SIET 2017
Submission Management System
Main Site
Submission Guide
Register
Login
Participant List
Abstract List
:: Abstract List ::

1 ABS-13 Business process management

Sentiment Analysis of Customer Engagement on Social Media in Transport Online
Melva Hermayanty Saragih, Abba Suganda Girsang

Bina Nusantara University


Abstract

Currently, social media in all of business organizations and community are commonly used to get many advantages such as feedback, customer engagement, and as tools of marketing and promotion. This research proposes investigating the customer engagement by analysis the comments on social media (facebook and twitter) in transport online. This study investigates by mining the comments of fan page facebook and tweet of twitter in three transport online in Indonesia, Gojek, Grab and Uber using the API service which is provided of both media social. The data comments are classified into some categories and positive and negative sentiment as well. These results show that the category ?Feedback system by driver? and ?Feedback system by user? have most comments for three transport online, while category ?Service quality for driver? has the smallest comments. This study also reveals the most comments are the complaining. This feedback of social media can be used to evaluate the performance of these business transport online.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=13


2 ABS-105 Business process management

Positive Impact of Customer Relationship Management (CRM) Implementation to Improving The Services of Animal Polyclinics Customers
Edy Prayitno, Novita Amylia Astuti

STMIK AKAKOM Yogyakarta


Abstract

Customer loyalty within the service organization is critical, so transactional data and interactions between customers and organizations must be well managed in Customer Relationship Management (CRM) to create an ideal relationship value.
Animal clinics need to use CRM to establish good relationships with their customers, animal owners. The purpose of this study is to know the positive impact of CRM applications on animal polyclinics.
CRM was built by involving five actors: the general public, animal owners, veterinarians, polyclinic leaders, and officers. Business process reengineering was done for several new services as a companys appreciation of customers. While other processes still follow the existing business process.
CRM was implemented by using a website that has various facilities that can be used by customers and polyclinics to more easily interact, and SMS (sort message service) service from polyclinics as a media attention of customers.
The applied CRM application can reduce service waiting time by 82% and add services by 50%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=105


3 ABS-102 Climate and ecosystem monitoring

Data Gathering in Rural Area using Publish/Subscribe over Wireless Mesh Network
Eko Sakti Pramukantoro, Kasyful Amron, Hanif Kuncahyo Adi

Universitas Brawijaya


Abstract

Observation data as a material in the decision making process becomes important. The difficulty of accessing data in remote areas poses a challenge to the development of the area itself. Wireless Mesh network infrastructure has proven to be applicable in rural areas. Publish/Subscribe mechanism also proves to be a communication pattern that can survive in unreliable situations. This research combines Wireless Mesh Network infrastructure and Publish/Subscribe mechanism to support the process of collecting sensor data in the rural area. Observed data will be collected on the local webserver. This data can later be transmitted with other technologies such as Delay Tolerant Network. To observe the data, the sensor takes more than 1 second with the percentage of success above 90%. In the data transmission process, not all data can be transmitted. There is a failure rate but no more than 20%. The number of nodes affects the delay in the data transmission process. Larger delays will occur in larger systems. Considering the ease of its application process in remote areas, its ability to survive in unstable network situations and relatively stable performance, this mechanism can be selected as an alternative data collection solution in the rural area.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=102


4 ABS-129 Climate and ecosystem monitoring

Drought Forecasting Using ANFIS on Tuban Regency, Indonesia
Andreas Nugroho Sihananto(a*), M. Shochibul Burhan (a), Arief Andy Soebroto (a), Wayan Firdaus Mahmudy(a), Fatwa Ramdani (a), Ahmad Luthfi (b), Hartanto (b)

(a)Faculty of Computer Science, Universitas Brawijaya, Malang, East Java, Indonesia
*andreas.nugroho90[at]gmail.com
(b)Meteorological, Climatological, and Geophysical Agency (BMKG), Karangploso Station, Malang, East Java, Indonesia


Abstract

Tuban is a regency on East Java Province, Indonesia, that always suffer from drought every year. A forecasting model is being needed to predict when the drought periods will happen next years. This paper studied the forecasting implementation by using two algorithms : Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) with implementation of additional weather parameters such as meridional wind and zonal wind that, as long as authors know, seldom to be used in weather forecasting model. Based on the RMSE test, ANN resulted RMSE 0.09145; 0.1288 ; and 0.1194 on three different regions while ANFIS resulted RMSE 0.01733; 0.01645; and 0.01714 on the exactly same three regions with ANN.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=129


5 ABS-184 Climate and ecosystem monitoring

Effect of Rotational Speed Spin Coating on pH Sensor Performance Based Titanium Dioxide (TiO2)
Eka Maulana, Onny Setyawati, Novvy Nurdiana Dewi

Department of Electrical Engineering
Brawijaya University


Abstract

The thick layer deposition of TiO2:AgCl on the alumina substrate by using spin coating method for pH sensor application has been done. Part of pH sensor in this investigation is electrode and sensing area, the electrode material is AgCl and the material were made to sensing area is TiO2. AgCl paste and TiO2 paste was made by mixing binder solution (PVA) with AgCl powder and TiO2 powder which stirrer with magnetic stirrer that the mixture became homogenous paste. The AgCl material were deposed to alumina substrate at various rotational speed spin coating 300 rpm, 500 rpm, 700 rpm, 1000 rpm, 1500 rpm, and the TiO2 material were deposed to alumina substrate at rotational speed spin coating 300 rpm. The result of sensor voltage showed that rotational speed spin coating effect on sensor output voltage. The sensor repeatibility result showed that was used of sensors repeatedly resulted in decreased sensor performance. This proves that the materials used make the sensor ineffective for reusable sensors. The sensitivity result of 1000 rpm and 1500 rpm sensors, showed that the optimum sensitivity is 1000 rpm sensor with -41,07 mV/pH. In the calculation of error analysis showed that the pH sensor has the best uncertainly (smallest) is 1500 rpm sensor that is 0,5484%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=184


6 ABS-197 Climate and ecosystem monitoring

Tide Level Forecast Using Grammatical Evolution
Nerfita Nikentari, Nola Ritha, Lidya Wati

Universitas Maritim Raja Ali Haji


Abstract

Tide level forecast is one vital part in maritime-related activities. A change in tide level can cause many effect in certain field such as flood risk, marine navigation, coastal structures, fisheries, and recreation. We applied Grammatical Evolution to forecast tide level. Experimental results showed that the applicability of Grammatical Evolution for forecasting long-term time series. Grammatical Evolution with best optimum function are able to predict tide level.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=197


7 ABS-1 Computer Engineering

Performance Testing Analysis on Web Application: Study Case Student Admission Web System
Mayang Anglingsari Putri, Hilman Nuril Hadi, Fatwa Ramdani

University of Brawijaya


Abstract

Websites used for universities selection entrance (admission) are most visited websites in daily activity, thus its performance is critical. The ability of web applications either to control or process users? requests determines its reliability. Furthermore, those websites which process students admission in Universitas Brawijaya and Politeknik Negeri Malang certainly engage huge volume of data and information that requires the highest level of reliability. Therefore, there is absolutely needed appropriate testing performances to measure the level of the certain application based on reliability rate. This measurement is used to determine responses, throughput, capability, and system scalability upon workload given. This research has a contribution to present testing performance concepts, goals, targets, types, and tools of Apache JMeter which is engaged for web assessment including detects mistake and error that relates to application performance and helps to improve the level of application performance as expected.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=1


8 ABS-7 Computer Engineering

Post-processing and Band Selection for Hyperspectral Image Data Classification with AdaBoost.MH
Desta Sandya Prasvita

Sekolah Tinggi Ilmu Manajemen dan Ilmu Komputer ESQ


Abstract

This research proposed a method to improve classification performance on hyperspectral image data. The method consists of three phases: 1) band selection, 2) hyperspectral image data classification, and 3) post-processing. AdaBoost.MH classifier is used for classification of hyperspectral image data. Accuracy of AdaBoost.MH classification without post-processing is 96.3%. After post-processing, accuracy increased by 3.698%. In addition to improving the accuracy of classification, post-processing also can remove the small misclassified pixel (speckle noise). To reduce the number of hyperspectral bands, band selection was used in this research. Band selection method used was contribution ratio value of AdaBoost.MH classifier. The maximum accuracy in this research is 99.9982%, only 1 pixel is misclassified.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=7


9 ABS-8 Computer Engineering

The Recognition of Mango Varieties Based on The Leaves Shape and Texture Using Backpropagation Neural Network Method
Fathorazi Nur Fajri, Nur Hamid, Ricardus Anggi Pramunendar

Sekolah Tinggi Teknologi Nurul Jadid, Sekolah Tinggi Teknologi Nurul Jadid, Universitas Dian Nuswantoro


Abstract

At this time, the demand of Indonesian mango is in great demand by the society especially for the superior quality mango like Manalagi and Gadung. However, many people are still wrong in distinguishing mango varieties. At this moment, the identification or introduction of mango varieties is done by eye. Some people may be expert in identifying mango varieties based on leaves by eye, but not all mango varieties they can identify. Until now, there are several methods to identify mango varieties, but the accuracy got is less than 80%. In research before, the extraction feature used is either shape or texture feature of the leaf images. In this research, we use Backpropagation Neural Network (BPNN) by using mango leaf shape and texture feature. A Dataset used are 300 images of mango leaves consisting of 150 images of mango leaves of Manalagi varieties and 150 images of leaves Gadung. By using this method, we obtain that the most optimal BPNN model got by using hidden layer = 19, learning rate = 0.9, momentum = 0.9, and epoch = 100 with root mean square error (RMSE) = 0.0018. The accuracy rate that we obtain is 96%

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=8


10 ABS-43 Computer Engineering

A Dual Dynamic Migration Policy for Island Model Genetic Algorithm
Alfian Akbar Gozali, Shigeru Fujimura

Waseda University


Abstract

The common problem in island model is the way to migrate individual from one to another island or usually called as migration policy. Previous researches in this movement protocol could be categorized into two different approaches, diversity preservation based such as migration protocol in LIMGA and better island pursuing based such as new dynamic migration policy. The main purpose of this works is to introduce a brand-new migration mechanism called as Dual Dynamic Migration Policy (DDMP) for island model GA. DDMP will take the advantage from result pursuer of dynamic migration policy and convergence avoider of LIMGAs migration protocol. The experiment result shows that DDMP could give great result while carrying out the general optimization cases. It could produce the best score result for all cases among previous island model migration methods. This work also compares DDMP with the best-known-so-far solution for the problem set.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=43


11 ABS-52 Computer Engineering

Design and Implementation 8 bit CPU Architecture on Logisims For Undergraduate Learning Support
Mochammad Hannats Hanafi Ichsan (a), Wijaya Kurniawan (b)

Computer Engineering
Faculty of Computer Science, University of Brawijaya
Jl. Veteran 08 Malang, East Java, Indonesia
a) hanas.hanafi[at]ub.ac.id
b) wjaykurnia[at]ub.ac.id


Abstract

Computer Architecture consist of some basic design based on Computer Organization and Architecture (COA) such as logic design. Leading students to improve knowledge about COA needs a comprehensive learning with a working simulation of a simple 8 bit Central Processing Unit (CPU). The Design based on Von Neumann Architecture that?s generally include Registers, Bus Interface, ALU, Memory and their structures. The CPU Architecture simulation used Logisims which capable to performs digital logic simulation. Logisims has an ability to perform digital logic to build subcircuits become a larger circuit in a single environment from low level combinational and sequential circuits to build a complete CPU. Undergraduate student has suffered to study COA with the theory that explained in several lesson, the lesson explained a structur and an architecture computer only. There are not providing a simulation that explained step by step computer architecture. This research to propose design and implementation 8 bit CPU architecture that designed and implemented step by step that created in Logisims. So the simulation easy to visualize then the Undergraduate Student can learn about the next competency which is to design and make a specific purpose computer.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=52


12 ABS-195 Computer Engineering

WEB Based Real Time Monioring Electrical Quantities Measurement
Dikpride Despa (a), Gigih F. Nama (b), Meizano A Muhammad (b), K. Anwar (a)

a) Department of Electrical Engineering
Engineering Faculty, Lampung of University
Jalan S. Brojonegoro No. 1 Bandar Lampung (3515), Indonesia
*despa[at]eng.unila.ac.id
b) Department of Informatics Engineering
Engineering Faculty, Lampung of University
Jalan S. Brojonegoro No. 1 Bandar Lampung (3515), Indonesia


Abstract

This study aims to build an applications, that can monitor the value of electrical quantities in a real-time using web technology at H Building of Engineering Faculty, University of Lampung (Unila). The whole system consists of several hardware and software components. Some hardware components such as current sensors, voltage sensors, signal conditioning circuit, embedded computer Arduino Uno, and Ethernet Shield. The programming language used for data acquisition was C language. The web-based applications produced the electrical quantities data measurement, appears in the form of text and graphical format that can be access anytime and anywhere. Some of web development tools used in this web development those are HTML, CSS, JS, JSON, and PHP programming languages. The results of monitoring data shown, the amount of electricity at distribution panels at this building was not balanced. This happen due the electrical load was unbalanced on each phase, other fact found that the trend of electricity usage still follows the trend of working hours.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=195


13 ABS-120 Conceptual modeling, languages and design

A Knowledge Base Repository Model for Multiple Domain Problem of Distributed Expert System Sevices
Istiadi (a*); Emma Budi Sulistiarini (a); Dedy Usman Effendi (a); Rudy Joegijantoro (b)

(a) Faculty of Engineering, Widyagama University of Malang, Indonesia,
* istiadi[at]widyagama.ac.id
(b) Department of Environmental Health Science, Widyagama Husada College of Health Malang, Indonesia


Abstract

An expert system that accommodates multiple domain problems is potentially accessed by a large number of users. This can degrade service performance if done at a single service center. The distributed service centers is an alternative to the problem, but the knowledge base between dispersed server needs synchronization mechanism. This study proposes a knowledge base repository model in the form of a database for multi-case and scattered expert systems. The repository model was developed to mark the knowledge base elements to be exchanged between servers. A data exchange mechanism is illustrated to accommodate the data exchange between expert system servers.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=120


14 ABS-147 Conceptual modeling, languages and design

Business Intelligence for Physical Examination Platform Service Reporting System
Abba Suganda Girsang, Sani Muhamad Isa, Aditya, Arie Purnama, Christopher Aryaguna, Evans Andita Sukmana, and Ferico Samuel

Bina Nusantara University


Abstract

As a human resource partner company which currently serves South East Asia area, this company is providing more than sixteen professional services to dozens of companies in South East Asia area. This Company is starting to provide physical examination platform service to its client since 2016 which effectiveness in choosing examination package and collaboration options are critical. Every decision made by the company could be disastrous due to the other competitors has been settled in the market for a long time. Therefore, capability to present reports or data in accurate and insightful ways is highly required in order for the company to make data-driven decision. The method chosen to develop the data warehouse along with its analytics and reports is Kimball methodology which has been introduced since mid-1980s and has been used by a lot of prior researchers. As the data can be displayed in various form as it needed, the stakeholder is able to make data-driven decision which benefit this company to perform better in the market

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=147


15 ABS-152 Conceptual modeling, languages and design

Integrated Social Media Knowledge Capture Model in Medical Domain of Indonesia
Farrell Yodihartomo (a), Dicky Prima Satya (a)

a) School of Electrical Engineering and Informatics, Bandung Institute of Technology
Jalan Ganesha 10, Bandung 40132, Indonesia


Abstract

The Social Media Platforms, as the one of largest part of today data traffic on the Internet, disseminate a vast volume of information, including medical information in it. Knowledge management system (KMS) approach is applied with a purpose to capture, maintain, and manage tacit or explicit knowledge available and collected within the social media platforms, organizations database, knowledge base, or document repository. Using Indonesian Natural Language Processing (InaNLP) and Data Mining approach, the proposed model is designed to improve the previous research related to social media knowledge capture model and offer a better accuracy, availability, and reliability of knowledge retrieved. Thus, the processed knowledge can be applied in medical sector of Indonesia, mainly aimed for medical practitioner to give a quick suggestion of the diseases regarding to the early diagnose which has been taken in the first place.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=152


16 ABS-160 Conceptual modeling, languages and design

MANAGEMENT COUNSELING SYSTEM FOR JUNIOR HIGH SCHOOL STUDENTS WITH KNOWLEDGE MANAGEMENT SYSTEM APPROACH
Novita, Agnes I.S., Faried, Isnin

Perbanas Institute - Jakarta


Abstract

Non-academic problems are often experienced by junior high school students that lead to, among others, disruption of academic achievements. Careful handling shall be necessary as the students at this level are still in their self-development stages and have yet found a good level of maturity. Thus far, school counseling system has been implemented but in reality the system has not been well documented by using an aid tool in information technology-based system.

A knowledge management system (KMS) is necessary in providing assistance to school counselors (?BK teachers?) in managing counseling processes, and by using this approach as the aid tool, thus non-academic problems experienced by the students can be dealt with.

This system is built by conducting a proper database design for manage the knowledge. The development process uses an SECI model approach and implements a tacit knowledge to explicit knowledge method in a knowledge management system that has been developed for purposes of counseling systems.

The knowledge management system that is developed has met with expected requirements in documenting the problems dealt with by the students as well as making BK teachers easier on making the best use of their own knowledge from experience in dealing with various problems that have been faced in their school.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=160


17 ABS-33 Data mining

Optimization of Fulfillment Nutritional Needs of Toddlers using PSO Algorithm with Flexible Budgeting
Leni Istikomah, Imam Cholissodin, Marji

Faculty of Computer Science Univeritas Brawijaya


Abstract

In the fulfillment of nutrients, one type of food alone is not enough so it requires a variety food ingredients that contain all the elements of nutrients. In this research give recommendation variation foodstuff automatically by using optimization process of Particle Swarm Optimization algorithm so that it can facilitate Posyandu and parents in providing daily food according to the nutritional needs of a toddler with economic conditions each parent. Based on the resulting in system accuracy up to 84.2% The expenditure of parents of children under five is 37.8%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=33


18 ABS-47 Data mining

A ALGORITHM HYBRID MODEL OF BAYESIAN FOR DETECTION PERFORMANCE IN UNHEALTHY LIFESTYLE
Ilham

UIN Sunan Ampel Surabaya


Abstract

This study was conducted to see the trend of diseases caused by unhealthy lifestyles in underserved communities and coastal villages around Gresik and Tuban by using a hybrid algorithm through the Bayesian Network structure construction.The purpose of this study is to produce a software model that is able to detect early disease risk propensity lagging rural and coastal communities who have unhealthy lifestyle in the form of construction of structures and produces a probability value with cendrungan disease.The validity of that people have a tendency to smoke will experience disease Tuberculosis, Bronchitis or Lung Cancer through the test system is 75% to 95%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=47


19 ABS-51 Data mining

Implementation Of Naive Bayes Classifier Algorithm On Social Media (Twitter) To The Teaching Of Indonesian Hate Speech
Naufal Riza Fatahillah (a), Pulut Suryati (a*), Cosmas Haryawan (a)

a) Department of Computer Information Systems, STMIK AKAKOM, Jl. Raya Janti 143,
Yogyakarta, Indonesia
lut_surya[at]akakom.ac.id


Abstract

Twiiter is a social media that is widely used as a medium of sharing on the internet. There are tweets containing sentences shared by the user, so they can be read by other users. A lot of information can be obtained from social media Twitter. Twitter users can connect with other Twitter users in international scale. Technology that is growing as today can be used for various things, especially in the information distributed in social media especially Twitter. One of the problems derived from social media is Twitter tweets containing speech are positive and negative utterances. From the above problems raised to be a research to classify tweets that contain positive speech and negative utterances using naive bayes classifier method. The results of this study are implemented into a system that can classify tweets on Twitter. The system is built using js Node technology and Naive Bayes classifier as the calculation method of classification. Based on the tests performed, the best accuracy is generated by systems using the Naive Bayes Classifier, which is 93%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=51


20 ABS-73 Data mining

Modeling Backpropagation Neural Network for Rainfall Prediction in Tengger East Java
Ida Wahyuni (a*), Nakhel Rifki Adam (b), Wayan Firdaus Mahmudy (a), Atiek Iriany (c)

(a)Faculty of Computer Science, Brawijaya University, Malang, Indonesia
ida.wahyuni8[at]gmail.com
(b) Faculty of Informatic Engineering, STMIK Asia, Malang, Indonesia
(c) Faculty of Mathematics and Natural Sciences, Brawijaya University, Malang, Indonesia


Abstract

Climate Change in the world gives many impacts on changing rainfall patterns. As a result, potato farming areas such as Tengger, East Java are having problems to determining potato growing times. It needs a method that can predict rainfall based on rainfall patterns that occur after climate change. Backpropagation Neural Network (BPNN) is one method that can learn from the data of the past and make it a benchmark to predict future data. In this research, BPNN method uses to predict rainfall in Tengger, East Java using rainfall data from 2005 to 2014 which is the period of climate change. Data from 2005-2009 use as training data and data from 2010-2014 use as data testing. In addition, this research also looking for the most optimal parameter modeling of BPNN includes the value of learning rate, hidden layer, and maximum epoch. Based on the test results, the most optimal parameter of BPNN is learning rate 0.2, hidden layer 3, maximum epoch 4000, and average error RMSE for 4 location in Tengger is 8.41. The result of RMSE error with BPNN is smaller compared to previous research using GSTAR SUR and Tsukamoto FIS.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=73


21 ABS-106 Data mining

Finding similar clustering pattern between students academic performance and non-curricular activities data
Nur Ayuni Ramadhani, Utomo Pujianto

State University of Malang


Abstract

Students attend a number of non-curricular activities in order to develop critical thinking skills and organizational skills. The negative side of this is that some students tend to be more active in the organizations activities than trying to improve their academic performance. This study focuses on cluster analysis on variables related to non-curricular organizational activities and student academic performance. The K-Means algorithm has been used to divide student academic performance into two clusters, high and low GPA. The results are then compared to the clustering of students - also with K-Means - based on the variable activity of each students organization. Experiments that have been conducted show that more than 50 percent of samples show an equivalent clustering pattern between each cluster of academic performance and their correlated non-curricular activity cluster.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=106


22 ABS-109 Data mining

Comparison of illiteracy cluster pattern and population data using Fuzzy C-Means.
Nimatul Rochmaniyah, Utomo Pujianto

State University of Malang


Abstract

Illiteracy is one of the problems for the Indonesian government that needs serious attention. A number of factors contributing to the high rate of illiteracy include unequal population spread and uneven distribution of teachers between urban and rural areas. One of the efforts made by the government to overcome the problem is by using a block system. However, the weakness of this system is that it takes a long time and also a huge cost. The solution provided to minimize the weakness of the system is to group regions with high and low illiteracy levels so as to form a group of data with the same characteristics. In this research, clustering process is done by using Fuzzy C-Means algorithm. The attributes used are the number of illiterate population, the total population, the number of poor people, the number of schools that have libraries, the number of community learning centers and the number of people who can not speak Indonesian. The experimental results have shown that these attributes have patterns that are identical to the pattern of cluster formation of illiterate population in an area.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=109


23 ABS-111 Data mining

Non-Linear Modelling of Variables Relationship in Multiple Time-Series Data with Extended Dynamic Interaction Network
Harya Widiputra, Elliana Gautama, Marsudi Kisworo

Perbanas Institute


Abstract

The topic of time-series modeling has been widely researched in studies of dynamic systems. Nevertheless, most studies have focused more on the task of modeling movement of a single time-series in order to forecast their future values and rarely gave further interest to learn about the governing behavior of the observed system. Therefore, we have previously introduced an adaptive machine learning method, named the Dynamic Interaction Network (DIN), to discover and model dynamic pattern of interactions from multiple time-series data to not only predict their future values but also to extract knowledge about their joint movement. However, the interactions were modeled only in a linear form which is a simplified representation of more complex relationships between variables related to real world phenomena. Accordingly, this research extends the previously developed method to allow the modeling of non-linear relationships between contributing variables in multiple time-series data. The objective is realized by incorporating the Extended Kalman Filter method into the DIN to enable the identification of non-linear interactions between variables in a set of multiple time-series data. Comparative study and results of conducted experiments reveals that the ability to model the dynamic interactions between variables in non-linear forms leads to better understanding of the nature of observed system and in addition helps to increase the prediction accuracy.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=111


24 ABS-122 Data mining

Optimization Fuzzy Neural Network with Simulated Annealing on Jatropha Curcas Plant Disease Identification
Diny Melsye Nurul Fajri, Triando Hamonangan Saragih, Andi Hamdianah

Brawijaya University


Abstract

Jatropha curcas are multifunctional plants, especially on the seeds. It can be extracted into oil as cosmetic
additives, detergents, waxes and biofuels. The Jatropha curcas farmer must be aware of the disease that
attacks by pest or virus for the existence and benefits of this plant. The main obstacle is the lack of
farmers?s knowledge about diseases that attack the plant, a system is needed that can utilize expert
knowledge so that it can give the decision like an expert. This paper used optimized of using simulated
annealing from Fuzzy Neural Network (FNN) method to identify Jatropha Curcas Disease and achieved
the best solution of 32.5%, better than just using FNN only.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=122


25 ABS-123 Data mining

Jatropha Curcas Disease Identification Using Fuzzy Neural Network
Triando Hamonangan Saragih, Diny Melsye Nurul Fajri, Andi Hamdianah

Brawijaya University


Abstract

Jatropha Curcas is a plant that has many functions and uses, but not apart from that this plant can be attacked by disease. Expert systems can be applied in identification so as to help both farmers and extension workers to identify disease. In this paper the method used in the identification of Fuzzy Neural Network (FNN) with an accuracy of 30% with an average accuracy of 11.2%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=123


26 ABS-130 Data mining

Gold Price Movement Forecasting using Hybrid FIS-ES
Andreas Nugroho Sihananto(a*), Fitra A. Bachtiar(a)

(a)Faculty of Computer Science, Universitas Brawijaya, Malang, Indonesia
*andreas.nugroho90[at]gmail.com


Abstract

Gold is the most popular commodities that has been used for exchange of goods for centuries. For a trader, the most important thing to gain profit on gold-exchange trading world is to know the behavior gold price, when it will be moving down or moving up in the stock market. Gold prices affected by some economic factors such as foreign exchange but most methods that studied gold forecasting only focused on technical analysis ? using time series and historical price. This paper is trying to formulate fundamental analysis to forecast gold price by using hybrid between Evolution Strategies and Fuzzy Inference System. It resulted lower error based on MAE compared on regular FIS, with the value is 9.460308 for regular FIS Mamdani and 2.73414 for ES-FIS.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=130


27 ABS-137 Data mining

The Performance of Genetic Algorithm Learning Vector Quantization 2 Neural Network on Identication of The Types of Attention Deficit Hyperactivity Disorder
Brillian Aristyo Rahadian (a*), Candra Dewi (b), Bayu Rahayudi (c)

a, b, c) Department of Informatics Engineering, Faculty of Computer Science, University of Brawijaya
*) brillianarc[at]gmail.com


Abstract

Attention Deficit Hyperactivity Disorder (ADHD) is a psychomotor disorder on children with characteristics like difficult to concentrating, difficult to keep quiet, feeling restless, unable to sit quietly and do something excessively. Detection types of ADHD is necessary to handle the patient appropriately. This research implements Genetic Algorithm Learning Vector Quantization 2 Neural Network (GA-LVQ2 NN). The advantages of LVQ2 are able to set the weight vectors on the process of supervised learning to estimate the classification results. But if the initialize weight vectors is not precise then the classification results will not be optimal. Genetic Algorithm (GA) is used to optimizing the weight vectors on LVQ2 training process. To find out the performance of GA for optimizing the weight vectors of LVQ2, then performed comparison between GA-LVQ2 and LVQ2. The testing is done by calculating the accuracy using 80 training data and 20 testing data. Based on the testing by 10 experiments, showed that the GA-LVQ2 method gives higher accuracy that is 89.5% compared to LVQ2 method which is 80%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=137


28 ABS-170 Data mining

Classification of Provinces Based on Schools Dropout Rate using C4.5 Algorithm
Annas Gading Pertiwi(a*), Triyanna Widyaningtyas (b), Utomo Pujianto (b)

Faculty of Engineering, University of Malang
Jalan Semarang 5, Malang 65145, Indonesia


Abstract

The education in a certain region reflects the region. In Indonesia, the rate of schools dropout rate in each province is considered high. This urges the effort to reduce the schools dropout rate. One of the ways to reduce it is by knowing the characteristic of province whose schools dropout rate is low. From those characteristics, we can classify to know which characteristic is similar to others, and it can be indicated as one of causative factors of the schools dropout rate itself. Data mining has been used to analyze data and can be used to solve this kind of problem. One of data minings method, C4.5 algorithm is chosen to be the method because its proved that the result is good and C4.5 is good for data which has many attributes. The result showed accuracy rate of 71.2%. Based on the result, it can be concluded that the development of the classification of provinces based on schools dropout rate using C4.5 algorithm was accurate to use.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=170


29 ABS-194 Data mining

Decision Tree to Analyze the Cardiotocogram Data for Fetal Distress Determination
Adhistya Erna Permanasari, Akhsin Nurlayli

Universitas Gadjah Mada


Abstract

Cardiotocography (CTG) is one of the most widely used techniques for recording changes in Fetal Heart Rate (FHR) and Uterine Contractions (UC). Assessing cardiotocography is crucial in identifying oxygen-deficient fetuses called hypoxia. This situation is defined as fetal distress and requires fetal intervention to prevent fetus death or other neurological disease caused by hypoxia. The proposed method in this paper is decision tree to analyze the Cardiotocogram data for Fetal Distress Determination. The main purpose of this classification method is to classify the fetal state class code consisting of normal, suspicious or pathologic. FHR pattern class or fetal state class code can be classified by pruned decision tree with the minimum misclassification error of classification confusion matrix. The misclassification error (0.184383) is the result of the experimental decision tree to analyze the cardiotogram data for fetal distress determination using pruned decision tree. By pruning the decision tree, 1593 + 130 + 11 = 1734 of 2126 is perfectly predicted, a promising result. 165 samples are predicted as ?suspicious?, and 138 as ?pathologic? whereas they have actual values of ?normal?. 62 ?normal? and 27 ?pathologic? classified instances have actual values of ?suspicious?. Additionally, 0 ?normal? and 0 ?suspicious? classified samples have actual values of ?pathologic?.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=194


30 ABS-17 e-Government and public sector

Environmental electronic governance (EEG) in Indonesia: its current conditions, challenges, and obstacles
Teguh Kurniawan

Public Administration Department, Faculty of Administrative Sciences, Universitas Indonesia


Abstract

Advances in information and communication technology have provided significant benefits in human life including in the field of environmental management. One form of ICT utilization in environmental management is through Environmental Electronic Governance (EEG). EEG according to Martin et.al (2012) is the rules, behavior and dynamic process by actors, including government, industries, market and civil society, using ICT to protect the environment and achieving sustainable development involved in environmental protection. As an instrument, EEG provides abundant tools for direct public participation, focuses on networking and interactions among stakeholders, and spurs citizens? interest in environmental problems and protection as well as facilitates their engagement. This paper tries to see how the current status of EEG in Indonesia and the challenges and obstacles it faces. Related to government actors, in this article the discussion focused on the Ministry of Environment and Forestry as the central actor and the Provincial Government of DKI Jakarta and the Municipal Government of Depok as regional actors.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=17


31 ABS-83 e-Government and public sector

Systematic Review of Critical Success Factors of E-Government: Definition and Improvement
Ruci Meiyanti (a*), Muhammad Mishbah (b), Darmawan Napitupulu (c), Dana Indra Sensuse (d), Yudho Giri Sucahyo (e)

a), b), c), d), e) Faculty of Computer Science, University of Indonesia, Jl. Margonda Raya, Beji, Pondok Cina, Kota Depok, Jawa Barat 16424
* ruci.meiyanti[at]ui.ac.id


Abstract

Critical Success Factors (CSFs) of e-government are useful to get success implementation of e-government services to the citizen. We collected CSFs of e-government from the various literature indexed by Scopus in the last five years with related topics. The problem isnt clearly meaning or too short word that revealed in the CSFs. The purpose of this study is to describe the CSFs? e-government. Describing CSFs? e-government is very important to avoid misinterpretation of the meaning. The clarification of CSFs? e-government will assist the developing concept in e-government of Indonesia. The approach of a method that used in this study is the systematic review using PRISMA 2009. The finding of this study is 52 the definition of CSFs? e-government from 39 paper and their influence to the stakeholders of e-government.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=83


32 ABS-84 e-Government and public sector

Influential Variables of Behavioral Intention to Use E-Government Services in Indonesia Using The Unified Theory of Acceptance and Use of Technology 2 Model
Ruci Meiyanti (a*), Deki Satria (b), Dana Indra Sensuse (c)

a), b), c) Faculty of Computer Science, University of Indonesia, Jl. Margonda Raya, Beji, Pondok Cina, Kota Depok, Jawa Barat 16424, Indonesia
* ruci.meiyanti[at]ui.ac.id


Abstract

Many studies about good services in e-government are done by improving services for the citizen. The behavioral intention to use e-government services can influence in improving services. There are variables used to measure behavioral intention to use e-government services. The research model used the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) model. UTAUT2 model is complete than UTAUT with added variables such as hedonic motivation, price value, and habit. In this study, we modified some variables in UTAUT2 such as price value variable was substituted by public value variable. Meanwhile, the hedonic motivation variable was eliminated because the activity of service was used to public service that was not to the business. Finally, we completed the model with antecedent trust variables because the intention to use in some research usually needs to trust. We had 203 respondents from the citizen that spread in Jakarta and surroundings to answer the online questionnaires about some e-government services in Indonesia. The findings of this research were three variables to give the positive impact on behavioral intention to use e-government services, they were public value, habit, and effort expectancy.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=84


33 ABS-19 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

DETERMINANTS OF E-COMMERCE SERVICE QUALITY, RECOVERY SERVICE QUALITY, AND SATISFACTION IN INDONESIA
Hendra, Ginting, Rini, Sembiring

Universitas Sumatera Utara, Politeknik Wilmar Bisnis Indonesia


Abstract

Managing recovery service quality, service quality and satisfaction of e-commerce services is highly significant businesses long-term growth. Previous research has shown that e-retailers experience difficulty maintaining customer satisfaction despite the recent rapid growth of Customer to Customer (C2C) e-commerce applications. Numerous studies have empirically examined mostly Business to Customer (B2C) e-commerce recovery service quality, service quality and satisfaction in various countires. Nevertheless, empirical research on these key contructs of C2C e-commerce in developing country like Indonesia is generally limited. Thus, the main objective of this paper is to identify whether recovery service quality and service quality have influence towards satisfaction in C2C e-commerce. This study draws on previous research to build a conceptual framework which hypothesizes relationships between these three e-commerce constructs and their antecedents. A survey was conducted among C2C e-commerce customers in the eastern province of Indonesia using a structured self-administered questionnaires. The results of this study show C2C e-commerce customer satisfaction in Indonesia is significantly influenced by service quality but is not influenced by recovery service quality, though when combined, it has has significant influence. The study limitations, implications, along with directions for further research are discussed.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=19


34 ABS-34 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

The Analysis Of Consumer?s Intention Model For Using E-Payment System In Indonesia
Sfenrianto, Junadi, Melva Hermayanty Saragih

Bina Nusantara University


Abstract

The objective of this research is to identify the factors that encourage consumers intention to use the electronic payment system for e-commerce transaction in Indonesia. This study was developed based on a model of factors influencing consumer?s intention to use e-payment, namely culture, perceived security, performance expectancy, effort expectancy and social influence. In conducting this study, the data is obtained from 110 respondents who used electronic transactions using e-payment. Meanwhile, the data collected was analyzed using Structural Equation Modeling (SEM). From the analysis, it is known that all of these factors significantly influence the value of the use of e-payment. The sequence is based on a significant level of the greatest to the smallest are effort expectancy, culture, social influence, performance expectancy and perceived security.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=34


35 ABS-50 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

An implementation of Delay-Disruption Tolerant Networking Approach in E-learning, Case Study: belajardisini.com
Eko Sakti Pramukantoro, Rakhmadhany Primananda

Faculty of Computer Science Brawijaya University


Abstract

Web-based applications can be accessed from anywhere at any time as long as the user is connected to the internet and works on a different platform. Besides for media information, website in educations is used for teaching and learning activities or known as e-learning. The benefits of using e-learning are student can learn anytime and anywhere, but when internet connection has a problem the benefits is not reached. In the implementation, e-learning highly depends on a stable internet connection. If there is a disruption in internet access, e-learning system is not working properly. Delay and disruption tolerant networking can overcome that problem. In this paper an e-learning system with DTN approach is proposed. We evaluate our system using several network condition and the result is a proposed e-learning system adaptive to disruption network and does not cause disadvantage when users working on the online quiz.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=50


36 ABS-71 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

Design of E-Commerce Chat Robot for Automatically Answering Customer Question
Adhitya Bhawiyuga, Moch. Ali Fauzi, Eko Sakti Pramukantoro, Widhi Yahya

Faculty of Computer Science
University of Brawijaya


Abstract

In order to provide an excellent service, the seller in e-commerce world is required to actively involves in communication with its customer. Nevertheless, in several condition e.g. in vacation or during a rest, the seller might be unable to communicate with his/her customer.
While employing a customer service can be a partial solution, it may involves additional costs for paying customer service persons.
In this paper, we propose the design and implementation of e-commerce chatbot system which provides an automatic response to the incoming customer-to-seller question. In general, the proposed system consists of two main agents : communication and intelligent part. In order to get the question message sent by the customer, the communication agent periodically performs a request to Telegram server using a standard HTTP protocol. Upon reception, it forwards that question to intelligent agent which then find the closest instance in predefined question-answer corpora. Notice that, we utilize the Levenstein distance to measure the difference between a submitted question with that of in predefined question-answer corpora. Once an closest instance is selected, the intelligent agent forward the answer to communication agent which then send the answer back to the sender through Telegram chat service. From usability and performance testing result, the proposed system can deliver the automatic answer in less than 5 seconds with relatively good matching accuracy.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=71


37 ABS-74 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

RELIABILITY TESTING USING HYBRID EXPLORATORY BASIS OF TOUR AND FUZZY INFERENCE SYSTEM TSUKAMOTO
Titis Sari Putri (a), Fatwa Ramdani (b)

(a) titis.sari.putri[at]gmail.com
(b) fatwaramdani[at]ub.ac.id
Faculty of Computer Science (FILKOM), Brawijaya University, Malang, Indonesia


Abstract

Reliability is a software quality attribute. Reliability can be used to indicate the software?s reliability. Reliability testing is an important matter in software development. It assesses the quality and reliability of a software system, especially a large-scale system that has plenty of users and handles many business operations and transactions. In this research, the reliability test was performed towards the two biggest Indonesian online marketplaces on the android platform, Bukalapak.com and Tokopedia.com. The testing applied scenario-based exploratory method which combines testing based on scenario and exploration. The test was known as hybrid exploratory testing. Scenario-based exploratory test was assessed to identify the result test that cannot be revealed by scenario testing, in which the performed testing is more accurate in resembling the actual end-user behavior that often differs greatly with the script in scenario testing. The testing results of Bukalapak and Tokopedia were compared and analyzed. The results showed that there were some weaknesses in the tested features: impurities, discomfort of using application, and potential occurrences of bugs or errors. From the scenario-based exploratory testing result, it can be concluded that there was no significant error that can damage the system. After optimization was conducted using Fuzzy Inference System Tsukamoto, it showed that both Bukalapak and Tokopedia have an equal end value output of the reliability system, though there were some different values of testing.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=74


38 ABS-144 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

The Acceptance of E-Payment among Indonesian Millennials
Bayu Kelana, Anggar Riskinanto, Deliar Rifda Hilmawan

STIMIK-ESQ


Abstract

The objective of this study is to explain how e-Payment system is being adopted by the millennial generation in Indonesia. This study applies the Technology Acceptance Model approach and a quantitative method, by analysing 424 PayTren e-Payment users, born between 1977 and 1998 using Smart PLS 3.0. The results show all independent variables have significant relationship to dependent variables. Among them, perceived ease of use has the strongest relationship with perceived of usefulness of e-Payment adoption. Meanwhile, the attitude towards using score shows the best predictive power in this model. The gender was proposed as a moderator variable but the findings did not support this role. Compared to e-Payment adoption in Jordan and Greece,?this study may provide a new perspective on how e-Payment being adopted by the millennial generation in Indonesia.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=144


39 ABS-145 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

E-Learning Readiness Measurement on Indonesian Student from Individual Perspective
Annisa Syatitah Muharina, Bayu Kelana

STIMIK - ESQ


Abstract

In 2010-2015, the e-Learning growth rates in Indonesia was above 30 percent, the top eight countries with the highest rates in the world. Therefore, the instrument of assessing e-Learning Readiness (ELR) in education institution in Indonesia is needed. By adapting Aydin & Tasci, Darab and Rohayati framework, this study develops the ELR instrument to see from individual perspective, in order to understand the readiness of Indonesian high school students with the e-Learning. The finding show that the student respondents are ready for e-Learning, but need some improvement particularly in two ELR areas, including innovation and self development. Although this instrument has been tested to measure high school students readiness, it can be adapted to be applied in different school settings in Indonesia.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=145


40 ABS-146 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

Tahani Model of Fuzzy Database for an Adaptive Metacognitive Scaffolding in Hypermedia Learning Environment (Case: Algorithm and Structure Data Course)
Akhsin Nurlayli (a*), Teguh Bharata Adji (a), Adhistya Erna Permanasari (a), Indriana Hidayah (a)

a) Department of Electrical Engineering and Information Technology, Universitas Gadjah Mada
Jalan Grafika 2, Yogyakarta 55281, Indonesia
*akhsin.nurlayli[at]mail.ugm.ac.id


Abstract

Studies show that an adaptive learning environment is needed by learners. Adaptive e-learning systems have been developed with instructions based on student[s] knowledge level, pedagogical recommender, and scaffolding using student[s] test scores. Since each student not only has different cognitive level but also has a different metacognitive level refers to student[s] knowledge and regulation, we currently develop an adaptive Hypermedia Learning Environment (HLE) for Algorithm and Data Structure Course based on metacognitive awareness level. HLE provides scaffolding as instructional interventions including pedagogical agents, guides, and strategies in learning to facilitate students[] aptitude in programming. Student[s] scaffolding can be determined by knowing his/her own metacognitive awareness level. In this paper, we propose the use of Tahani Model of Fuzzy Database in web-based HLE to support the adaptive decision making. Tahani is one of the feasible models in fuzzy database that can categorize the students[] metacognitive awareness level using Metacognitive Awareness Inventory (MAI). The produced fuzzy rules have been validated by metacognitive experts of educational psychology. A hundred students were categorized into three groups according to three metacognitive awareness level (low, medium, high) and getting their own scaffolding. The results verified the suitability of the proposed model and the student[s] scaffolding.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=146


41 ABS-148 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

Data Warehouse Development for Hotel Reservation system
Abba Suganda Girsang, Sani Muhamad Isa, Bambang Susilo, MaxLian, Danang Satya, Salman Al Fariz, Dudi Ramdani

Bina Nusantara University


Abstract

XYZ company is a major travel company that has many branches in Indonesia. As the company always provide adequate services for their customers, the company grows larger and need the fast report for business development in order to compete with it?s fellow rivals. By creating a data warehouse for XYZ company, this paper aims to exploit business possibilities stored in XYZ company to get better information about customer behavior from the data received from each branches to get information for the company to take the decision. The Data Warehouse development is done through the nine step methodology designed by Kimball and Ross. Furthermore, the data is able to be presented in dashboard or report corresponding to the user to simplify the data presentation, The data warehouse is used to integrate the data sources needed to provide quick and accurate information.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=148


42 ABS-169 e-Healthcare, e-Learning, e-Manufacturing, e-Commerce

SPOC Adoption in Accounting Course among Indonesian Undergraduate Students
Anggar Riskinanto, Bayu Kelana, Indah Navidah Hayati

STIMIK ESQ


Abstract

This research aims to determine the factors that affect behavioral intention of Indonesian university students to use Small Private Online Course (SPOC) for accounting learning. This research applies the Unified Theory of Acceptance and Use of Technology 2 (UTAUT 2) approach and a quantitative method, by analysing 42 undergraduate students in accounting course in an Indonesian university (STIMIK ESQ), using Smart PLS 3.0. The finding show that hedonic motivation has a significant effect to behavioral intention of SPOC adoption in accounting learning. The gender and experience were proposed as a moderator variables but the findings did not support this role. This may provide a new perspective on how SPOC being adopted in accounting course by undergraduate students in Indonesia.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=169


43 ABS-117 Energy independent

NEW AND RENEWABLE ENERGY : A REVIEW AND PERSPECTIVES
Deria Pravitasari, Sapto Nisworo

Universitas Tidar


Abstract

The issue of global warming and climate change caused by the greenhouse effect becomes one of the interesting things to be studied. This is done to encourage the use of environmentally friendly energy to reduce greenhouse gas emissions resulting from the use of fossil fuels. Government with various policies to be a leading player in the development and application of environmentally friendly energy technology. Reviews and perspectives on renewable energy will be presented in this paper. Based on the review, more research, education and support is needed to further promote and apply technology engineering to process existing resources and seek new renewable energy opportunities to meet policy-defined objectives and standards.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=117


44 ABS-118 Energy independent

Potential of Irrigation Channels as New Renewable Energy Sources
Sapto Nisworo, Deria Pravitasari

Universitas Tidar


Abstract

The energy potential of the Progo Manggis Channel is designed and functioned as an irrigation channel, so that the flow discharge is relatively stable throughout the period, even in the dry season. The geographical location of the city of Magelang is relatively ups and downs, to avoid the erosion of the design of Progo Manggis Channel is designed to occur in many dams, it is possible to investigate the energy potential that is found throughout the dams. The potential energy produced is very possible to be used as a cheap alternative energy and environmentally friendly. Measurement of discharge is done by measuring the flow velocity and cross-sectional area of the channel. Measurement of discharge with the aid of current flow meter through velocity-area method approach with total calculation of 510,4303 kW, Kedungsari 61,63076 kW, Menowo 2 55,20043 kW, Menowo 1 114,647 kW, Polosari 80, 67755 kW, Kebonpolo 110,40009 kW, Rindam 21,35771 kW, Jl. Iklas I 25,81356 kW and Jl. Iklas II 40.70273 kW. If at two points of the dump with the highest energy value will be built small-scale generator, then in accordance with the Ministerial Decree No ESDM. 1122 / MEM / 2002 on the Small Scale Power Generation Manual, which states that PT PLN must purchase at 60% of the Cost of Generation in the area to be connected to a low voltage network and 80% connected to a MV network

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=118


45 ABS-185 Energy independent

Internet-Based Monitoring and Protection on PV Smart Grid System
Sholeh Hadi Pramono, Eka Maulana, Sapriesty Nainy Sari

Department of Electrical Engineering
Brawijaya University
Jl. MT Haryono 165 Malang


Abstract

Smart Grid system based on PV development is an effort to increase the efficiency of energy use. Hence the need for tools that can perform protection and monitoring of the use of electric power should be manage optimally. This tool is able to monitor the current and voltage on the load by using current and voltage sensors, then the data is converted into power (P). This system is also capable of performing power protection to the load so that if the power supplied by the power supply exceeds the power specifications on the burden of this protection system will disconnect the load from the power supply. In this study the method used in monitoring and protection mechanism using wireless communication via internet with ESP 8266 Wifi module which can be accessed through website. This tool is capable of reading up to 5 A current changes with an error readings on ammeters by 2.4% and the change in voltage up to 30 V with an error in the voltmeter readings of 0.073%. The data read on the microcontroller will be stored on the memory card such as data loggers and some are sent to the web service is available to do the monitoring. The protection system in the tool uses the relay, the relay has a switching response at speeds of 54 ns. This system can be used for the control and protection mechanisms of future smart grid networks based on IoT (Internet of Things).

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=185


46 ABS-16 ERP and supply chain management

Garbage Transportation Route Basedon Vehicle Capacity and Container Condition in Central Surabaya
Ricky Setiawan Hartanto, Edwin Hendrawan

Petra Christian University


Abstract

The garbage collection constitutes the main contributor in terms of the garbage cost management. Garbage collection scheduling model is a variable that determine cost center. Cities in developing countries still operate a traditional waste transport and handling where rubbish were collected at regular intervals by specialized trucks from curb-side collection or transfer point prior to transport them to a final dump site. The waste generated by Central Surabaya is about 190 tons / day spread over 18 TPS. Garbage transport from TPS to Benowo TPA uses 2 types of trucks. The first truck is an ordinary truck located in Tanjung Sari Sanitation and Plantation Service and has a carrying capacity of 10,939 Kg. The second truck is a compactor truck located in the Menur Sanitation and Plantation and has a carrying capacity of 10,300 Kg. Effective working time of DKP Surabaya is 8 hours / day. The optimization of transport is done by making the transportation route. Optimization is done by considering the technical aspects of the mass of waste in each TPS and truck capacity. Routes using the shortest distance, not the fastest time. Ritivity is adjusted according to the garbage mass of each TPS. The optimal route for regular TPS is 12 routes with a total distance of 580 Km, while the optimal route for TPS compactor is 9 routes with a total distance of 505 Km. The garbage mass of each TPS and vehicle capacity has a significant influence on the waste transport route.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=16


47 ABS-89 Human computer interaction

User experience measurement on virtual dressing room of madura batik clothes
Ari Kusumaningsih, Arik Kurniawati, Cucun Very Angkoso, Eko Mulyanto Yuniarno, Mochammad Hariadi

Informatics Departement, University of Trunojoyo Madura, Electrical Department, ITS Surabaya, Indonesia


Abstract

The exoticism of Madura batik has become one of the tourist charm of the Indonesian Madura island. Madura batik created by skilled hands with a high level of artistic imagination to produce diverse and unique batik. Our virtual dressing room use augmented reality technology to create solution of an efficient shopping experience by superimposing the 3D model of a Madura batik cloth within the live Kinect sensors. The superimposed 3D model of Madura batik will then transformed using 3D rigid transformation and bond to the movements of customer so it appears as if the customer is wearing the virtual dress in the live video view. We built the system using SimpleOpenNi library, and Microsoft Kinect SDK. User Experience (UX) measurement on the adaptation key performance indicators which include Attention, Importance, and Arousal showed 96% of respondents said that they were satisfied while the Quality key performance indicators which include Impression, Valence, and Enjoy showed 97% respondents were satisfied.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=89


48 ABS-15 Image processing and pattern recognition

Android Based Application for Recognition of Indonesian Restaurant Menus Using Convolution Neural Network
Windra Swastika, Yoko, Hendry Setiawan, Mochamad Subianto

Fakultas Sains dan Teknologi, Universitas Ma Chung


Abstract

Indonesia has a lot of typical food which is rich in flavor. However, compared with countries such as China, Thailand or India, Indonesian food is not widely known by the international community. One of the factors that cause Indonesian food is less popular is that the food information can not be accessed easily by foreign travelers during a visit to Indonesia. Food name is written on the restaurant menu can be read, but not really showing the information about the food, such as ingredients, how the food is prepared or a picture about the food itself. Hence, it is required to facilitate foreign travelers who want to find out information about Indonesian food based on restaurant menus. An Android based application has been developed to show food information. The application captures a text on the restaurant menus, process and recognize the text using convolution neural network (CNN). The recognized text is then matched with predetermined database to show the information about the food. The application was able to recognize 100% of the menus when the menus uses Sans Serif Font. However, the accuracy dropped into 56% when the menus uses Times New Roman.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=15


49 ABS-29 Image processing and pattern recognition

Chicken Meat Freshness Identification using the Histogram Color Feature
Rosa Andrie Asmara (a*), Faisal Rahutomo (b*) and Qonitatul Hasanah(c*)

a) Malang State Polytechnic, Information Technology Department, Malang, Indonesia
*rosa.andrie[at]polinema.ac.id
b) Malang State Polytechnic, Information Technology Department, Malang, Indonesia
*faisal.polinema[at]gmail.com
c) Malang State Polytechnic, Electrical Engineering Department, Malang, Indonesia
*qonitatul[at]polinema.ac.id


Abstract

Identifying the chicken meat freshness level is necessary since it involves the quality of the meat consumed. This research aims at identifying the freshness level of chicken meat based on the color histogram feature. The color histogram feature that used is RGB. RGB value is acquired from the image sample of chicken breast meat. First Order Statistical Method is used to reduce the color feature dimension such as Mean, Max, Min, and Standard Deviation. The value is then classified using Na?ve Bayes Classifier, Support Vector Machines (SVM) Classifier and C4.5 Decision Tree. The Classification method then compared for analyzing the best classifier. The freshness of chicken meat level defined in three class, fresh, medium, and old. The chicken meat classified as fresh from 0 to 4 hours been slaughtered, 4-6 hours classified as medium, and more than 6 hours classified as old. The result of color histogram feature by Na?ve Bayes method shows 38.24%, Support Vector Machine (SVM) shows 50%, whereas C4.5 decision tree method shows 70.59% classification accuracy. The classification process of the chicken meat?s freshness level based on the color histogram feature suggests using C4.5 method which

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=29


50 ABS-36 Image processing and pattern recognition

Colour image segmentation for malaria parasites detection using cascading method
Yonathan Ferry Hendrawan, Cucun Very Angkoso, Rima Tri Wahyuningrum

Informatics Engineering
University of Trunojoyo Madura


Abstract

Malaria disease is one of the deadliest diseases. According to WHO report, in 2016, there were 212 million malaria cases with 429,000 estimated deaths around the world. Rapid and accurate detection of malarial infection could certainly help in reducing the above numbers. This paper focuses on segmentation and detection of microscopy images infected with malaria parasites using colour-based cascading method. The method begins with RGB normalization process, followed by gamma correction, then noise reduction, exposure compensation, edge enhancement, fuzzy c-means clustering, and lastly, morphological processes. We used the method to detect malaria infection for four plasmodium types and three malaria development phases in 574 images. The experimental results demonstrated that the method achieved 97.92%, 98.60%, and 98.26% for sensitivity, specificity, and accuracy . These results suggest that the proposed method is able to detect malaria in microscopy images robustly.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=36


51 ABS-60 Image processing and pattern recognition

Retinal Blood Vessel Segmentation Using Multi-Scale Line Operator and K-Means for Data Preprocessing
Winda Cahyaningrum(a). Randy Cahya Wihandika(a). Agus Wahyu Widodo(a)

(a) Faculty of Computer Science, University of Brawijaya Malang


Abstract

The vascular changes in retinal blood vessel are precursor of a disease, such as diabetic retinopathy, stroke and hypertension. Changes can be seen by analyzing retinal image manually, but it takes a long time. In this research we propose the automation of vascular segmentation processes in retinal image so that can assist in analysis process, which is an important step in retinal image analysis. The segmentation process is done by detecting the linear structures using the multi-scale line operator and preprocessing the images using k-means. Linear structure detection using single-scale line operator is performed on several different scales, then the results of each scale are combined. Image preprocessing using k-means algorithm aims to remove the optic disc area, which will be detected as false positive. The k-means algorithm is employed to cluster three parts in the retinal image, which are the background, foreground, and vessel. The performance of the proposed method was evaluated using the DRIVE and STARE datasets. The segmentation result was observed visually and shows that the K-Means algorithm can remove false positive and preserve true positive around the optic disc area. The evaluation using the DRIVE dataset shows the average accuracy and area under curve of 0.9421299 and 0.78004, respectively, whereas it reaches the accuracy and area under curve of 0.949293361 and 0.778, respectively, using the STARE dataset.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=60


52 ABS-67 Image processing and pattern recognition

Light Traffic Sign Recognition using Edge Detection and Eigen-face
Hurriyatul Fitriyah, Edita Rosana Widasari, Gembong Edhi Setyawan

Departement of Computer Engineering, University of Brawijaya


Abstract

Most traffic sign recognition algorithms utilize Template Matching which compare detected sign with templates. Studies on this method have shown outstanding recognition accuracy. Nevertheless, the Template Matching burdens a system in term of memory usage since it has to store numerous templates. Eigen-face is a basic method originated to recognize faces. It is efficient and practical since system only needs to store an Eigenface-image and Weights that associated with it. This paper developed a traffic sign recognition using Eigen-Face algorithm. Instead of using RGB images, the learning was utilized edges. It is more distinctive feature compare to color intensity which vary from yellow, red and blue and additional black symbol. The template signs were first converted into grayscale intensity. Its edges were detected using common Sobel approximation and then concatenated into one matrix. Eigenvalues and Eigenvectors of the matrix?s Covariance were then calculated. In this algorithm, the biggest Eigenvector was selected and projected as Eigenface-image. Each traffic sign had Weight associated with the Eigenface-image. This paper compare how disperse and distinct each sign?s weights with and without color pre-classification based on median of Hue. The recognition with color pre-classification shown clearer weights? distinction between each type of traffic sign yet lower weights? disparity within types.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=67


53 ABS-79 Image processing and pattern recognition

On The Usage of Hybrid 1-D Convolutional Network and Long- Short-Term- Memory Network for Multiple-Site Fatigue Damage Prediction on Aircraft Lap Joints
Muhammad Ihsan Mas, Timotius Devin, Lintang Adyuta Sutawika, Mohamad Ivan Fanany

Universitas Indonesia, Faculty of Computer Science


Abstract

Multiple site fatigue damage is a problem that affects many operators of aging aircraft. The methods currently in place for prediction of such damage are conservative and sensitive to noise and cannot fully account for grain-level material variations, which results in aircraft being more conservatively designed than they need to be. The authors augmented the dataset of the FAA AR-07/22 report into a sizable body of variable- and constant-amplitude multiple-site fatigue damage sequences. This was done implementing and tuning the existing Walker crack growth model, implementing the plastic zone linkup criteria for crack interaction, as well as adding Gaussian noise at different stages of the computation. The interim model used for predicting the damage is a hybrid 1-D convolution and bi-directional LSTM model, which achieved an average of 70.7% accuracy, 110.3% MAPE, and 6.5% MSE on the interim version of the dataset. A detailed breakdown of the error characteristics and the hyperparameters that have salient effects on the performance of the model are also examined.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=79


54 ABS-90 Image processing and pattern recognition

Multiplication of V and Cb Color Channel Using Otsu Thresholding for Tomato Maturity Clustering
Yuita Arum Sari, Sigit Adinugroho, Putra Pandu Adikara, Abidatul Izzah

Computer Vision Research Group, Faculty of Computer Science, Brawijaya University.

Politeknik Kediri


Abstract

One of necessary stages of doing digital image processing is laid on the preprocessing phase, in which one of those techniques is segmentation. Segmentation plays important role of separating a tomato as object and background. When capturing tomato image at outdoor, the segmentation for each device can be different and causes lower performance of maturity clustering problem. This paper proposes new framework in to enhance the evaluation measure of clustering by using combined segmentation of V and Cb color Channel. V color channel is taken from YUV color space, while Cb is obtained from YCbCr color space. Both two color channels are segmented using Otsu thresholding since Otsu is robust to segment each tomato image. Color feature extraction is then applied to the retrieved color of segmented image. In this paper, RGB and L*a*b* color space model is used to get the feature, but only R, G, a*, b* color channel are examined. The 6-Means clustering algorithm is also applied to agglomerate tomato maturity based on six color levels. There are four smartphone cameras to capture the tomato image with unconditional lighting condition at outdoor. The experimental result shows that the smallest value from the average of Mean Square Error (MSE) in all devices reach 3.135. This indicates that the new framework is able to cluster the tomato maturity with only use G and B color channel.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=90


55 ABS-100 Image processing and pattern recognition

Feature Selection with Genetic Algorithm for Alcoholic detection using electroencephalography
Rivaldy Septyasurya, Indah A. Siradjuddin, and Arif Muntasa

Informatics Department
University of Trunojoyo Madura
Jl. Raya Telang
Kamal, Bangkalan, Jawa Timur 69162


Abstract

Electroencephalography (EEG) is a technique of recording human brain waves through electrodes mounted on the human scalp. The obtained signal data can be used as a detector of alcoholic criteria. Currently, the used method to detect an alcoholic through a urine test or a breathalyzer test, and it takes time and hassle. Therefore, the usage of EEG for alcoholic detection is proposed in this research. There are four stages in this research. First, noise removal of the recorded signal using Independent Component Analysis. Second, feature extraction with Discrete Wavelet Transform. Third, extracted features from the previous stage are selected based on the importance of each feature using Genetic Algorithm, and final stage is the alcoholic detection with Learning Vector Quantization. Experiments were conducted on 64 channels of EEG data, and the feature selection stage using Genetic Algorithm makes the accuracy of the detection process is increased. The average accuracy of the detection is 74.6%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=100


56 ABS-114 Image processing and pattern recognition

Automatic Arrhythmia Identification Based on Electrocardiogram Data Using Hybrid of Support Vector Machine and Genetic Algorithm
Reiza Adi Cahya (a), Candra Dewi (b), Bayu Rahayudi (c)

a, b, c) Faculty of Computer Science, Universitas Brawijaya
Jalan Veteran, Malang 65145, Indonesia


Abstract

Electrocardiogram (ECG) recordings provide insights on a person?s cardiac activity, and can be used to identify heartbeat abnormalities, or arrhythmia, they might suffer. Automatic ECG interpretation can be achieved via machine learning techniques to aid physicians. This research aims to model a ECG classifier based on Support Vector Machine (SVM) and Genetic Algorithm (GA) to classify a normal beat and three types of arrhythmia. SVM with radial basis function (RBF) kernel were used because of its superiority in handling the large amount of numerical features generated from the ECG. GA was used to enhance the SVM by performing feature selection to generate dataset with optimal amount of features, limiting possibilities of feature redundancy. ECG dataset used for model training and testing were obtained from the Massachusetts Institute of Technology?Beth Israel Hospital (MIT?BIH) arrhythmia database. The GA-SVM classifier then was compared to SVM-only classifier to ensure that GA is indeed able to improve the capabilities of SVM. Results show that in classifying six-seconds ECG recording with 120 training data dan 20 testing data, GA-SVM yielded better average accuracy of 82.5%, compared to 47% yielded by SVM-only classifier.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=114


57 ABS-121 Image processing and pattern recognition

A Review Paper on Attendance Marking System based on Face Recognition
Khem Puthea, Rudy Hartanto and Risanuri Hidayat

Universitas Gajah Mada


Abstract

Attendance marking system has been become a challenging, intriguing and accurately in the real-time system. It is tough to mark the attendance of a student in the large classroom, and there are many students attend the class. Many attendance management systems have been implemented in the current research. However, the attendance management system by using facial recognition still has issues which allow the research to improve the current research to make the attendance management system working well. The paper will do a literature review on the previous worked from different researcher has done on their research paper. This paper does not only provide the literature review on the earlier work or related work, but it also provide the deep analysis of Principal Component Analysis , discussion, suggestion for future work.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=121


58 ABS-134 Image processing and pattern recognition

Indonesian Food Items Labeling for Tourism Information Using Convolution Neural Network
Renaldi Primaswara Prasetya; Fitra A. Bachtiar

Faculty of Computer Science Brawijaya University


Abstract

Abstract-The process of recognition and classification of food through digital image processing system has developed and widely applied, especially to know the nutrients or food ingredients contained. This system is certainly useful if applied to help foreign tourists who are in Indonesia who want to know information about Indonesian food, considering they are not accustomed to the type of food in Indonesia. The types of food in Indonesia vary diverse in each city area but have similar physical appearance to each other. In this research, we will apply Convolution Neural Network method to classify Indonesian food type and extract the name or information about the food. By using CNN method, obtained the results of the food classification which is quite reliable reached 77% accompanied by information in accordance with the type of food.
I. INTRODUCTION AND BACKGROUND
The process of recognition and classification of food by using computer vision method becomes quite interesting research as done by [1] - [4], food classification process can be used to know information about nutrition or material contained in the food, even also used to know calorie content [5] in food. It is often used for someone who is on the diet or for someone who is in the process of health restoration [6], and many other applications. Some difficulties in recognition and classification of food is the diversity of food types especially food in Indonesia
Type of food in each region or city in Indonesia, has its own characteristics and is very diverse, but in appearance has an almost indistinguishable similarity. Sometimes it make own difficulties for foreign tourists who travel to Indonesia who want to know the type of food information, and the content of food in Indonesia, considering the tourists are not accustomed to eating Indonesian food. Even, Indonesian citizens may not be able to recognize a type of food that is a typical food in a particular area.
Getting information manually of direct submission is quite time consuming and needs to remember it, even we can forget it, while information through browsing cannot be done if we do not know the type of food before. Therefore, in this research we will do the process of classification on Indonesian food which will give information about the name of food and food content by using CNN method which have ability in process of deep learning and processing big data of food which is diverse.
The process of recognition is conduct by inputting the food image as an input that will be recognized directly by the system. Initially, input image will undergo a training process first through feature extraction with feedforward and backpropagation approaches by involving the convolution process as well as the looping process to reduce the spatial size of the image. The results of the classification of food names that have been recognized by the system will then be translated through information obtained from the wikipedia website. Convolution Neural Network generally has three main layer components used in the classification process including convolution layer, layer pooling and activation function. In this research will be used max pooling function at layer pooling, while in phase of activation function will be used activation function of ReLU.
Convolutional Neural Network (CNN) is the development of Multilayer Perceptron (MLP) designed to process two-dimensional data. CNN is a reliable method of deep learning to solve complex problems and requires large data sets [2], so it is expected to be able utilized to classify Indonesian food types that are quite diverse but have similarities with each other.
II. RESULTS
To know the performance of Convolution Neural Network method in recognizing some types of Indonesian food, we do some experiments and testing involving 5 types of food and performs 45 times experiments. This experiment implemented on Intel Core i5-5200U processor 2.7GHz using python 3.6.1. Training dataset image of the Indonesian food was taken from website using fatkun batch downloader which consist of 350 ? 500 images on each type of Indonesian The dataset of food used is limited to traditional Javanese food such as rujak cingur, gudeg, soup tuna, soup soto, and gado-gado. The results of classification and information about some Indonesian foods that have been recognized by the system are shown as follows:

III. CONCLUSION
The convolution neural network method is quite good and accurate in recognizing several types of Indonesian food that has similarities with each other. This is evidenced by the accuracy that reached 77% by using 500 times the iteration process. Some deficiency in this research are, the system is less able to distinguish between buntut soup and soto soup. In addition, this method has a deficiency in distinguishing between positive image (food data) and negative image (non-food data), so it still allows unnecessary recognition process when it detects an object that is not actually a food.
For future work, it can consider the number of iterations and learning rate although the computation process will run longer. In addition, further research can also utilize other recognition methods such as Haar Cascade method so it could reduce the level of error in the process of recognition.

REFERENCES

[1] C. Paper, ?Highly Accurate Food / Non-Food Image Classification Based on a Deep Convolutional Neural Network,? no. September 2015, 2016.
[2] P. Napoletano, G. Ciocca, P. N. B, and R. Schettini, ?Food Recognition and Leftover Estimation for Daily Diet Monitoring Food Recognition and Leftover Estimation,? no. November, 2015.
[3] M. Ogawa, ?Comparative Study of the Routine Daily Usability of FoodLog : A Smartphone-based Food Recording Tool Assisted by Image Retrieval Journal of Diabetes,? no. February, 2014.
[4] L. Herranz, L. Herranz, S. Jiang, S. Member, and R. Xu, ?Modeling Restaurant Context for Food Recognition Modeling restaurant context for food recognition,? no. October 2016, 2017.
[5] H. Kagaya, ?Food Detection and Recognition Using Convolutional Neural Network,? no. 2.
[6] L. Notes and C. Science, ?FooDD : Food Detection Dataset for Calorie Measurement Using Food Images FooDD : Food Detection Dataset for Calorie,? no. January 2016, 2015.
[7] Hubel, D. and Wiesel, T. (1968). Receptive fields and functional architecture of monkey striate cortex. Journal of Physiology (London), 195, 215?243.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=134


59 ABS-150 Image processing and pattern recognition

Improving Performance Batik Image Classification based on Bag of Visual Word using Ensemble Classifier
Mulaab, Achmad Jauhari

Department of Informatics
Madura Trunojoyo University


Abstract

Automatic feature descriptor from images is an important role in the textural image retrieval and classification. In particular, batik image classification is essential to preserve the wealth of traditional art of Indonesia. Image batik has its unique pattern characteristic such as color intensity, ornament visualisation and ornament size. In order to recognise any patterns of batik need feature extraction methods. The scale invariant feature transform (SIFT ) can be used as an image feature descriptor in some applications. This paper presents an efficient based on Bag of Visual Words (BoVW) with features of scale invariant feature transform (SIFT) and implemented ensemble classifier for improving classification accuracy. The result of the experiment obtained the average of classification accuracy is 0.95

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=150


60 ABS-156 Image processing and pattern recognition

Arabic Letter Segmentation using Modified Connected Component Labeling
Fakhry Ikhsan Firdaus (a), Achmad Khumaini (b), Fitri Utaminingrum (c)

a,b,c ) Faculty of Computer Science, Brawijaya University, Indonesia


Abstract

The Arabic letters are very different from the Latin. Arabic letters can be interconnected with other letters, in addition, the Arabic letters also have many variety of shapes based on its position. The important point to obtain high accuracy for recognizing Latin Arabic letter is the segmentation process. This is a challenge when we will segment each letter. Especially if the letter is interconnected. Many researchers have proposed methods for segmenting Arabic writing in the previous research. In this paper, we propose Modified Connected Component Labeling (MCCL) to perform Arabic letter segmentation. Performance of Connected Component Labeling (CCL) in the previous research to separate each interconnected object still needs an improving. For under certain conditions there are still letters that are connected with other letters so that the letters are still not segmented correctly. To solve that problem, we apply MCCL with multiple rules based on moment location, object location, and height of the object to segment each letter. We evaluate our proposed method using quantitative analysis compared to CCL method. Our method successfully segmented each Arabic letter. Referring to the testing results, the performance of MCCL is better than CCL.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=156


61 ABS-161 Image processing and pattern recognition

Smoker?s Detection System based Tongue Reflectance Analysis in Hyperspectral Imaging
Iqbal Fachrizal, Bramma Kiswanjaya, Adhi Harmoko Saputro

Department of Physics, Universitas Indonesia, Indonesia
Department of Dentistry, Universitas Indonesia, Indonesia


Abstract

Smoker and non-smoker tongue is hard to differentiate visually, even by an experienced doctor or dentist. One of the most objective ways to recognize the smoker tongue is by using tools such as a camera. In this paper, we proposed a system that contains two parts, hardware, and software. The hardware consists of the workbench, camera slider, mini studio box, two halogen light source and hyperspectral camera with a spectral range between 400-1000 nm connected to PC via camera link. The system complemented with hyperspectral image processing software built up especially to analyze the smoker tongue. The reflectance values of tongue surface was extracted from respondent tongue image that previously corrected using white and dark hyperspectral image references. Averaging all of the spectral data have been done to transform the existing features into a lower dimensional space. The principal component analysis (PCA) was used to compute and select the features subset which will be used as an input to the classifier. The support vector machine (SVM) classifier is used as image classification model since it performs excellently to choose the best hyperplane separator between two different classes. Fifty hyperspectral images was classified as normal tongue and smoker?s tongue which examined and analyzed by the system. Filled up questionnaires cross-checked evaluation of the system results and interviewing respondents by a dentist to validate the accuracy of the system analysis. The result of the tongue analyzing the system to identify the smoker is satisfying with its accuracy.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=161


62 ABS-164 Image processing and pattern recognition

Text Detection and Recognition using Multiple Phase Method On Various Product Label for Visual Impaired People
Rizdania (a), Fitri Utaminingrum (a)

(a) Faculty of Computer Science (FILKOM), Brawijaya University, Jalan Veteran 8, Malang, East Java, Indonesia


Abstract

This paper reviews one of a noteworthy objective in research range of computer vision and digital image processing, which are text detection and recognition. Text detection and recognition is one of the popular research fields. The result of the study is useful for the visual impaired people, because it may help them in buying and choosing their preferred product in the market. This study applied Multiple Phase (MP) method for the text detection process and the text recognition. The text detection was going through some phases which are the Maximally Stable Extremal Regions (MSER) detection, canny edge detection, region filtering, and Optical Character Recognition (OCR).The OCR was used for text recognition process. The experimental result for our proposed method performance was 80,88%, which was better compared to the previous research which used the two-stages classifier that was only 69% of performance.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=164


63 ABS-171 Image processing and pattern recognition

Prediction of Soluble Solid Contents Mapping on Averrhoa Carambola using Hyperspectral Imaging
Maisyarah Yuniar Rangkuti, Cuk Imawan, and Adhi Harmoko Saputro

Departement of Physics, Universitas Indonesia


Abstract

In this research, hyperspectral imaging system has been developed to determine the quality of fruits based on the profile mapping of soluble solid content (SSC) in Averrhoa carambola with combining spectral and spatial analysis. The proposed system consists of a Specim FX-10 Hyperspectral Camera with spectral range 400-1000 nm, workbench slider, two 150 Watt halogen lamps tungsten fitted with an angle of 45? to illuminate in the camera?s field of view and a personal computer. A push-broom technique is applied to acquire hyperspectral images from all sample in the visible and near infrared region of 400-1000 nm. The region of interest (ROI) of each sample is obtained at 15 ? 15 pixels from the center position of each sample. All of the samples were analyzed using partial least squares (PLS) to obtain prediction models for SSC mapping on star fruit. The developed model is then used to create the distribution mapping of the soluble solids content on the starfruit. The performance of the prediction model was evaluated by observing the correlation coefficient, and root means square error. The number of samples used is 250 samples. The prediction model of PLS result provided correlation coefficient of 0.91 and root mean square errors of 0.50. The overall performance showed that hyperspectral imaging system might be useful to predict and to map soluble solid content of star fruit and suitable in an industrial sorting fruit system.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=171


64 ABS-172 Image processing and pattern recognition

Prediction System of Bruising Depth of Guava using Visual-NIR Imaging
Ida Ratna Nila, Cuk Imawan, and Adhi Harmoko Saputro

Departement of Physics, Universitas Indonesia


Abstract

This paper presents a proposed system to detect bruising depth on guava during its shelf-life time based Visual NIR image. The hardware of system consists of the workbench equipped with the hyperspectral camera and two 150 W halogen lamp. The proposed image processing technique perform hypercube image correction, feature extraction, feature selection and prediction model. The averaging value of spatial distribution was used as feature extraction of the local area of the bruise part. The partial least square was used to select the best feature that has an appropriate correlation with the target. The multivariate regression was performed to develop a model of the bruising depth prediction system. A total of 173 guava samples were used for analysis and divided into two groups (unbruised and bruise). Bruises were manually induced using steel ball with a constant height to simulate a damage caused by mechanical injury. Samples were analyzed in 1st, 3rd, 4th, 5th, and 6th days. The first 35 guavas was used as a sample for the calibration model, whereas the remaining 138 were used for the prediction model. Relationships between shelf-life time and bruise depth were investigated. Finally, the optimised model was used to generate the distribution map of bruising depth in the guava samples under various shelf-life time after bruising. The result was found to be helpful for the application in the industry to disquish the bruise and unbruised guava.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=172


65 ABS-175 Image processing and pattern recognition

Prediction of Paddy Field Area based on Aerial Photography using Multispectral Camera
Prasepvianto Estu Broto, Adhi Harmoko Saputro, Dony Kushardono

Department of Physics, Universitas Indonesia, Indonesia
Indonesian National Institute of Aeronautics and Space (LAPAN), Indonesia


Abstract

Calculation of paddy field area is necessary to know to predict a number of the paddy production in some region. In this paper, we proposed a system to predict paddy field area automatically based on aerial photography that collected using the multispectral camera. The multispectral camera is mounted on the wing of an integrated aircraft with GPS to provide position information of the resulting image. The multispectral images from the camera are then processed using texture segmentation. The threshold value is obtained by using Otsu method to convert the multispectral image into a binary image. Ground Sampling Distance (GDS) is computed by knowing the height of the aircraft flying and the field of view of the camera. GSD values can be used to calculate the area by multiplying by the number of pixels of each class. The system evaluation is performed by comparing the results of software processing with manual processing and calculated by confusion matrix. The accuracy of the system that has been made is 98%. This research was succeeded to make an automatic prediction system of paddy field area based on aerial remote sensing which the output is information of paddy field area.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=175


66 ABS-143 Information infrastructure for smart living spaces

IoT based Electrical Energy Consumption Monitoring System Protoype : Case Study in G4 Building Universitas Negeri Malang
Dyah Lestari, Irawan Dwi Wahyono, Irham Fadlika

Department of Electrical Engineering Universitas Negeri Malang
Jalan Semarang 5 Malang 65145
Indonesia


Abstract

Energy monitoring system becomes an important subject to provide information of electricity usage for the users. Moreover, with rapid development in information technology, especially IoT, it is possible to establish better energy monitoring system by providing real-time consumption data. In this paper, IoT based Electrical Energy Consumption Monitoring System Protoype for G4 Building Universitas Negeri Malang is developed. Real-time measurement of the energy consumption utilizes current and voltage sensors for each wiring phase of the building electrical panel. For the IoT system, data is processed and displayed in Web based system using Public Subscribe method. This prototype is implemented during work-hours and achieving 88.90% accuracy based on the electrical data of the G4 Building

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=143


67 ABS-22 Information retrievel

Tourism Destination Rating System Based on Social Media Analysis (Proposal and dataset development)
Diana Mayangsari Ramadhani (a*), Faisal Rahutomo (b), Cahya Rahmad (b)

a) State Polytechnic of Malang Electrical Engineering Department Malang, Indonesia
*diana_mayangsari[at]polinema.ac.id
b) State Polytechnic of Malang Information Technology
Department Malang, Indonesia


Abstract

It is very important for tourists to get information about their next adventure. Many websites offer those tips, one of them is Tripadvisor which has a destination rating feature that makes it easier. However, the rating score is managed manually by filling a scoring page available on each tourism location. This paper collect datasets for tourism destination automatic rating system based on social media analysis. The first dataset is a category and sub-category of adjective, which helps the semantic analysis process. The second dataset is a manual review of the tourist destination which is done by 11 people, added with the review from Tripadvisor. Results from the second database are used as the supporting data for the rate.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=22


68 ABS-85 Information retrievel

Focused Web Crawler for Indonesian Recipe
Gusti Ahmad Fanshuri Alfarisy and Fitra A. Bachtiar

Faculty of Computer Science,Universitas Brawijaya, Malang, Indonesia


Abstract

Crawlers are commonly used to traverse and store all public webs that are connected through links. However, the conventional crawlers could not be used for crawling or storing web pages with a particular topic. Thus, in order to store a web page in the case of Indonesian recipes, focused crawlers proposed. This paper, propose focused web crawler for Indonesian food recipes using simple classification of the web page based on the analysis of Indonesian recipes available on the internet, providing priority levels of a link through anchor text and URLs, and restricting the traverse by the depth. The results show that the proposed crawler can collect recipes on the web based on user query.

I. INTRODUCTION AND BACKGROUND


Today, a lot of the data available on the internet is generated by the internet users themselves that can be accessed publicly. For instance such as Facebook posts, video content on Youtube or image on the Instagram that impact on tremendous velocity and volume of data available on the internet. This case also applied to recipes. Not only written by the bloggers or official website, but also written by the internet users openly.

In Indonesia, all recipes on the internet are not well organized that associated with the unavailability of the search engine that focused on Indonesian recipes. The most widely used search engines are Google and Bing which are designed to index all web pages that do not focus on certain topics like recipes.

Therefore, if all the Indonesian recipes available on the internet can be collected, it will provide many advantages. This collected data can be exploited more widely such as diets for chronic diseases, calorie estimation, classification, clustering, information extraction, recipe search engine, or the generation of the new recipes based on a collection of recipes.

In order to collect web pages, a web robot called web crawler is used that became the basis of collecting the web before indexing is performed by search engines [1]. However, since web page volume is extremely large, it is very difficult for the web crawler to explore all area. Even for the general search engine, it can?t reach all the web pages available on the internet [2]. The crawler method that commonly used and found in web crawler library is using breath first and depth first [3]. Furthermore, it is highly likely that spesific query which is relevant is not indexed in search engines. Thus, the objective of the web crawler can be directed to specific need which is traversing deeply the web that relevant to the topic [4]. This specific objective of web crawler called as focused web crawler that search, collect, index, and maintain as much relevant web as possible on the query and avoid collect or traverse the irrelevant web [5].

Therefore, this study emphasizes the application of focused web crawler devoted to Indonesian recipes. Firstly, the user query is determined and the seed URL is retrieved from a third party search engine that is already mature. Simple classification techniques are used that provide good results on a characteristic of recipes. This classification technique is based on the analysis of 100 different recipes that include food, beverage, and snack. The next visit also determined based on the priority value obtained from the similarity of anchor text and URL to the query through Jaccard similarity. The depth of visits is also given with certain threshold to minimize irrelevant web visits and handling tunneling phenomena in a focused web crawler. By using this technique, the focused web crawler can be applied to Indonesian recipes.

II. RESULTS

The testing is performed on four different query for 100 web page for each query. It is intended to test how robust the crawler using the different keyword that includes food, beverage, and snack. The keyword is chosen as follows :

1. Resep Ayam Goreng (Fried Chicken Recipe)
2. Resep Brownies (Brownies Recipe)
3. Resep Ikan Nila (Tilapia Recipe)
4. Resep Teh Susu (Milk Tea Recipe)

The results show that the proposed method can download the relevant web page on Indonesian recipes. Whereas by using crawler that uses breath first search, a lot of irrelevant web pages are downloaded even using good seed URL.

III. CONCLUSION

The proposed method for focused web crawler for Indonesian recipe is comprised of recipe classification and the priority level of the visit. In classifying whether a page is a recipe or not, a simple technique is used that used the index of the word in a web page. The priority value is given based on the value of similarity between keywords with URL and anchor text. The depth of the visit is also limited to minimize irrelevant visits.

As a result of the proposed method, 327 of the 400 downloaded web pages are relevant to the recipe topic. This shows that the focused crawler with this technique can be used to download recipes for Indonesian cuisine based on the query of the user. As a result of the proposed method, 327 of the 400 downloaded web pages are relevant to the recipe topic. This shows that the focused crawler with this technique can be used to download recipes for Indonesian cuisine based on the query of the user. In the future study, term extraction for the relevant page may be used to increase the efficiency of crawler when a lot of irrelevant pages are visited.

REFERENCES

[1] S. Brin and L. Page, ?Reprint of: The anatomy of a large-scale hypertextual web search engine,? Comput. Networks, vol. 56, no. 18, pp. 3825?3833, 2012.
[2] L. Jayaratne and S. Lanka, ?A STATE-OF-THE-ART SURVEY : FOCUSED WEB CRAWLING USING NAMED ENTITY RECOGNITION FOR,? vol. 4, no. 1, pp. 45?54, 2014.
[3] ?A collection of awesome web crawler,spider and resources in different languages.? [Online]. Available: https://github.com/BruceDone/awesome-crawler. [Accessed: 16-Aug-2017].
[4] A. Gupta and P. Anand, ?Focused web crawlers and its approaches,? 2015 1st Int. Conf. Futur. Trends Comput. Anal. Knowl. Manag. ABLAZE 2015, pp. 619?622, 2015.
[5] S. Chakrabarti, M. Berg, and B. Dom, ?Focused crawling: A New Approach to Topic- Specific Web Resource Discovery,? Comput. Networks, vol. 31, pp. 1623?1640, 1999.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=85


69 ABS-97 Information retrievel

IMPROVING CLASSIFICATION PERFORMANCE OF PUBLIC COMPLAINTS WITH TF-IGM WEIGHTING
Fakhris Khusnu Reza Mahfud, Aris Tjahyanto

Department of Information System, Institut Teknologi Sepuluh Nopember


Abstract

Currently Media Center e-Wadul still uses manual labeling in the process of complaint submission, as a result Media Center administration takes a longer time in coordinating with regional work unit (SKPD) to respond to complaints registered. Therefore, it is necessary to classify complaints based on SKPD to speed up the timing of complaint submission. The challenge of the study is that data consists of many classes that are more than two classes. The proposed method of this research is TF-IGM weighting that can overcome multiclass problems. TF-IGM can calculate distinguishing class precisely of a term. The famous Term Frequency-Inverse-Document Frequency (TF-IDF) and TF-Binary weighting methods are also used as a comparison. The classification is performed on Support Vector Machine (SVM), Naive Bayes and K-Nearest Neighbor (KNN) algorithm. In this research, the incoming public complaints will be processed through the pre-process stage, feature selection stage, term weighting stage, and classification stage. Accuracy, precision, recall and f-measure values were used in this study to evaluate classification performance.

Index Terms : E-Government, Media Center E-Wadul, Text Mining, Text Classification, TF-IGM
INTRODUCTION
Media Center E-wadul is one of E-Government application which is used by Surabaya City Government to accommodate peoples participation in complaint, information and suggestion on urban development process implemented by Surabaya City Government. Complaints, information or suggestions coming into the media center e-wadul will go through several processes. In the first process, the incoming information will be submitted by the E-Wadul media center administrator to the relevant regional work unit (SKPD) in the Surabaya City Government through the Public Complaints Service Team (TPKPM). After the admin gets answers from the SKPD then the admin responds to complaints on the media center E-Wadul within 1x24 hours. The problem lies in the process of submitting complaints from the admin to SKPD because the labeling of complaints is done manually based on SKPD that handles the complaint. As a result, SKPD and administrator should take a long time to respond the public complaints on media center E-wadul. To solve the problem, this research proposes automatic text classification.
Text classification or text categorization is the task of assigning one or more than one categories to the documents in a collection form a set of known categories [1]. The stages of the text classification consist of feature extraction, feature selection and classification [1]. Several methods of classification are often used in processing the text i.e SVM [2][3][1][4], Naive bayes [4][1] and KNN [2]. The challenge of the study is that data consists of many classes that are more than two classes. The appropriate weighting and can represent the document can have a direct impact on improving the classification results [2].
In previous studies the classification was used in the Thai restaurant review [5], agricultural documents [3], twitter social networks [4], news texts and academic abstracts [6]. Other studies suggest that the Term Frequency Inverse Gravity Moment (TF-IGM) can precisely calculate the class distinguishing power of a word and take advantage of the inter-class word distribution [2]. TF-IDF was used as a comparison in this study as it proved effective in previous studies [2].
The method proposed in this study is the classification because the complaint data already has a label in accordance with the SKPD. Naive Bayes, KNN and SVM classification methods were used in this study. TF-IGM was used as a term weighting in this study, TF-IDF and TF-binary was also performed as a comparison. TF-IDF is used because it has been proven to be an effective scheme in information retrieval and other text mining tasks [2].
METHOD
In this research the method used is data preparation method, preprocess method, feature selection method and classification method.

Data Preparation
Data collected from March 2017 until April 2017 with data collection method using R. Package software used is "Rvest", then the data has been collected will be stored in .csv format.
Preprocess
After going through the data preparation stage, then proceed to the pre-process stage. In the first step, case folding. Case folding, tokenizing, filtering, stemming.
Term Weighting
Term weighting by the TF-IGM, TF-IDF and TF-binary method. The purpose of this weighting is to look for terms that can represent a document by converting it into a vector form. Excess TF-IDF is efficient, easy and has accurate results [7]. The advantages of TF-IGM are able to precisely determine the strength of the distinguishing class of a term[2].
Classification Method
Classification method that will be used in this research is SVM, Naive Bayes and KNN.
SVM is a classifier that has the purpose to find the function of the separator (hyperplane) with the largest margin, it can separate the two data sets optimally. SVM is very effective and fast to solve text data problem. Text data matches the SVM algorithm because text tends to have a high dimension [8].
Naive Bayes is a simple probabilistic classifier based on the Bayes theorem. This classifier can be very efficient and accurate especially when the number of variables is high [8].
KNN is an algorithm that works by comparing the distance of input data with a number of k training data closest [9]. This algorithm is robust to overcoming noisy training data [10].
Performance calculation
Calculation of performance in this research is accuracy, precision, recall, and F-measure
Accuracy is the value of the comparison between the values of data that are correctly classified with all data [4]. Precision is the ratio between the amount of data in a class that is correctly classified with all data in the same class [4]. Recall is the comparison between the amount of data in a class that is correctly classified with all data classified in the same class [4]. F-measure is a single parameter of the size of a retrieval success that combines recall and precision [4].
REFERENCES
[1] A. Rehman, K. Javed, and H. A. Babri, ?Feature selection based on a normalized difference measure for text classification,? Inf. Process. Manag., vol. 53, no. 2, pp. 473?489, Mar. 2017.
[2] K. Chen, Z. Zhang, J. Long, and H. Zhang, ?Turning from TF-IDF to TF-IGM for term weighting in text classification,? Expert Syst. Appl., vol. 66, pp. 245?260, Dec. 2016.
[3] A. D. Putri, ?Klasifikasi Dokumen Teks Menggunakan Metode Support Vector Machine dengan Pemilihan Fitur Chi-Square.,? 2013.
[4] A. A. Arifiyanti, ?Tesis - EKSTRAKSI FITUR PADA KONTEN JEJARING SOSIAL TWITTER BERBAHASA INDONESIA DALAM PENINGKATAN KINERJA KLASIFIKASI.? 2015.
[5] N. Claypo and S. Jaiyen, ?Opinion mining for Thai restaurant reviews using neural networks and mRMR feature selection,? in Computer Science and Engineering Conference (ICSEC), 2014 International, 2014, pp. 394?397.
[6] A. Hamzah, ?Klasifikasi teks dengan na?ve bayes classifier (nbc) untuk pengelompokan teks berita dan abstract akademis,? in Prosiding Seminar Nasional, 2012.
[7] S. Robertson, ?Understanding inverse document frequency: on theoretical arguments for IDF,? J. Doc., vol. 60, no. 5, pp. 503?520, Oct. 2004.
[8] C. MEGAWATI, ?ANALISIS ASPIRASI DAN PENGADUAN DI SITUS LAPOR! DENGAN MENGGUNAKAN TEXT MINING,? 2015.
[9] A. D. Arifin, I. Arieshanti, and A. Z. Arifin, ?Implementasi Algoritma K-Nearest Neighbor Yang Berdasarkan One Pass Clustering Untuk Kategorisasi Teks,? 2012.
[10] F. P. Shah and V. Patel, ?A review on feature selection and feature extraction for text classification,? in Wireless Communications, Signal Processing and Networking (WiSPNET), International Conference on, 2016, pp. 2264?2268.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=97


70 ABS-99 Information retrievel

Increased Information Retrieval Capabilities on e-Commerce Websites using Scrapping Techniques
Deborah Kurniawati, Deny Triawan

STMIK AKAKOM Yogyakarta


Abstract

When the number of accessible sources of information is so much and even unlimited, the information retrieval process becomes very complicated. Visiting information resources one by one and comparing data or information from all the information sources visited will add much time to the process of rediscovering the information. It takes a technique that can gather information from multiple sources into a single entity to facilitate the process of information retrieval.
This study uses 3 e-commerce websites as a source of information. By utilizing the crawling procedure will generate new variables that can store data from the source information, which then these data will be stored in a database. Web crawling works by taking html tags as needed, using scrapping techniques.
Furthermore, by collecting the data in the database, the information retrieval process can be done easily, and by using the query this process can be done in less time with 100% precision.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=99


71 ABS-193 Information retrievel

Clusster Testing for Music Genre Based On Active Frequency Using K-Means Method
M. Syahrul Munir1), Hardianto Wibowo2)

Universitas Pembangunan Nasional Veteran Jawa Timur1)
Universitas Muhammadiyah Malang2)


Abstract

In everyday life we find many genres or types of music around us but not all genres want to be heard by music lovers because of several factors such as the problem of taste or even because of the mood or atmosphere of the listener who allows to hear one type of music. Categorization is required in order to group the different types of music according to genre or type. By detecting every frequency of each music we can know how soft or aggressive the music is to the listener, this is the underlying research to reduce the burden of users in grouping various types of music. By using the software tooldari unity the author conducted a study to detect the frequency movement of the types of music so that it can determine clusters by using the K-Means method. The results of the calculations can be used as a reference to see how similar or different the types of music dangdut, EDM, metal, pop / rock, reggae and acoustics based on active frequency.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=193


72 ABS-57 Intelligent transportation systems

An Initial Framework of Hybrid Evolutionary Algorithm (EA) with Multiple Criteria Decision Making (MCDM): Plant Forecasting
Januardi Nasir, Azizul Azhar Bin Ramli

1.Universitas Putera Batam,Student PhD Universiti Tun Hussein Malaysia, 2.DEPUTY DEAN UNIVERSITI TUN HUSSEIN ONN MALAYSIA | PARIT RAJA 86400 BATU PAHAT JOHOR DARUL TAKZIM


Abstract

Crop forecasting is very important for the needs of dragon fruit planting in the tropics as uncertain climate can make it difficult for farmers to determine the appropriate time to plant. That is why farmers need logical information. To find a suitable time to planting. This research framework aims to build a system for forecasting the plant season based on rainfall humidity and wind direction by combining the evolution algorithm (EA) which has four algorithms that can be used, one of which Grammatical Evolution (GE) Will be combined with one of the fuzzy methods of multiple criteria decision makings. A decision-making method that aims to establish an alternative forecasting of a number of alternatives based on certain criteria so that in the incorporation of this algorithm will have a good adaptive level so that it can get a prediction model.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=57


73 ABS-72 Intelligent transportation systems

LOCATION DETERMINATION OF CITY TRANSPORT SHELTER BASED ON THE PREDICTION OF PASSENGER OCUPANCY ON ROUTE OPTIMIZATION SYSTEM (Studi Case : City Transport in Bandung)
Sri Suryani Prasetiyowati, Yuliant Sibaroni, Mahmud Imrona

School of Computing, Telkom University


Abstract

Transportation is a very strategic research area to be developed and it is related to population mobility. The problems associated with transportation are complex, such as capacity and facilities of roads, the type and number of transportation modes, and also the behavior of passengers and the drivers, and the regulations in the field of transportation. One of the modes of public transportation is city transport. This mode of transportation is very interesting to discuss, because it has some unique characteristics, such as the number of mode or route of city transport that increases over time, the behavior of the city transport drivers when pickup or drop the passengers in any place. Solutions to overcome the attitude of the undisciplined driver and passengers is done by forcing the driver to drop and pick up the passengers on the provided place. The determination of the location of shelters at this particular location will be the focus of this study where the determination of the shelter is based on the high passenger occupancy rate based on the Kriging method. The result obtained is a system that can display the location points, which is the proposed a shelters for city transport, with the average passenger occupancy approaching the required passenger capacity.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=72


74 ABS-31 IT management and governance

Evaluation of Information Technology Implementation on Business Goal Improve And Maintain Business Process Functionality at Economic Development Group East Java Representative Office
Awalludiyah Ambarwati, Ariinta Deraya Ratulangi

Program Studi Sistem Informasi Fakultas Ilmu Komputer
Universitas Narotama
Surabaya, Indonesia


Abstract

Economic Development Group East Java Representative Office is one of the groups/divisions in a institution that uses the most information system application in every business process. This group activity includes collecting and processing raw data from survey into final data, calculating and determining the inflation rate of East Java, as well as empowering Micro, Small and Medium Enterprises (MSMEs) that can drive economic growth. As a Representative Office in East Java, all Information Technology/Information System (IT/IS) implementations refer to the Information Systems Management Department located at the headquarters. However, this group able to develop applications as needed to support their activities. This research was conducted to evaluate IT/IS implementation in order to ensure alignment between IT goals and the organizations business objectives. This assessment taken from internal perspective views of Balanced Scorecard focusing on business goal Improve And Maintain Business Process Functionality using Control Objectives for Information and Related Technology (COBIT) 4.1. There are two domains and several IT Process used to measure maturity level and gap analysis. Domain Plan and Organise (PO) consist of two IT Process namely PO2 and PO3. While domain Acquire and Implement (AI) consist of six IT process, there are AI1, AI2, AI4, AI5, AI6, and AI7. The research results obtained existing maturity level value in Economic Development Group East Java Representative is 3.64. This value is categorized as managed and measured level. It indicates this group had an integrated information system which could support the business process.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=31


75 ABS-37 IT management and governance

Modified Genetic Algorithm for High School Time-Table Scheduling with Fuzzy Time Window
Ruth Ema Febrita (a), Wayan Firdaus Mahmudy (a)

(a) Faculty of Computer Science Brawijaya University, Jalan Veteran No.8, Malang 65146, Indonesia


Abstract

Time-table scheduling in educational institution is a very complex problem due to many regulations that must be considered referred to hard and soft constraint. Crowded schedulling in high schools level can make students too tired and distracted during the learning process. Time-table sceduling which offer the right time window is the purpose of this research, so the students will have their best time to study every subject, which requires different thinking portion. We proposed a modified genetic algorithm with fuzzy time-window to solve this problem. The proposed method has been successed to find optimum solution for this problem, with the best pop size set at 210, crossover rate=0.7, mutation rate=0.3 and maximum iteration set at 80 generations.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=37


76 ABS-192 IT management and governance

Heuristic Evaluation and ISO 9126 in Implementation of Decision Support Sytem for Selection of Outstanding Marketing Officer BRI Katamso Yogyakarta
Ripto Mukti Wibowo 1,2) , Adhistya Erna Permanasari 2), Indriana Hidayah 2)

1). Blitar State Community College, Indonesia
2). Departement of Electrical Engineering and Information Technology. Faculty of Engineering,University of Gadjah Mada, INDONESIA


Abstract

Decision Support System (DSS) computerized using the Multi-Criteria Decision Making is expected to help the decision maker to provide alternative decisions, while the final decision is determined by the decision maker. After the development process of the system of Election of marketing Officer (MO) conducted the evaluation process. With an evaluation of the designed system is expected to provide answers to existing problems and improve goal attainment. Evaluation used to evaluate DSS is using heuristic evaluation and ISO 9126. ISO 9126 to determine the level of effectiveness, efficiency, satisfaction. Heuristic Evaluation system to help evaluate User Interface Selection of DSS MO. Based on the ISO and heuristic evaluation, the general design of the interface and some of the factors effectiveness, efficiency, satisfaction resulted in the average value of 3.75 and the DSS can be used by the BRI Katamso but there are still some things that need to be improved, especially in ease of use.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=192


77 ABS-59 Media, game and mobile technologies

Mobile App for Stock Prediction Using Improved Multiple Linear Regression
Abidatul Izzah (a), Yuita Arum Sari, (b) Ratna Widyastuti (c)

a) Informatics Engineering, Polytechnic of Kediri
Jln. Mayor Bismo 27 Kediri, Indonesia
abidatul.izzah90[at]poltek-kediri.ac.id
b) Computer Science, Brawijaya University
Jln. Veteran No.8, Malang, Indonesia
yuita[at]ub.ac.id
c) Informatics Engineering, Polytechnic of Kediri
Jln. Mayor Bismo 27 Kediri, Indonesia
ratna.widya[at]poltek-kediri.ac.id


Abstract

Stock Prediction is developed in both of two studies, economics and data mining. Stock predictions got special attention because it is important for create a more effective and efficient planning. So that, in this study, we built a mobile application based android platform to predict stock prices using Improved Multiple Linear Regression (IMLR). IMLR is a hybrid Multiple Linear Regression with Moving Average technique. The app is built in several steps, they are requirements analysis, system design, implementation, and testing. Data is collected from the finance.yahoo.com page with category "Jakarta Composite Index (^ JKSE)" which is automatic taken by using Yahoo Finance API. In this app, users can see daily stock history also stock price predictions in real time.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=59


78 ABS-101 Media, game and mobile technologies

Metamorphosys edugame using 2048 game rules
Arik Kurniawati, Fahrur Rozi, Yonathan Ferry Hendrawan

University of Trunojoyo Madura


Abstract

A game can have many derivatives encompassing diverse topics. One of those games is 2048 game which originally is all about adding numbers. This paper describes how we built a game that introduces its player to metamorphosis process using 2048 game rules on a mobile platform. The experimental results showed that teenagers and adults could play the game well, but not so for children.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=101


79 ABS-136 Media, game and mobile technologies

Optimized Walking Straight Guidance System for Visually Impaired Person That Use Android Smartphone
Rosa Andrie Asmara; Fais Al Huda; Cahya Rahmad; Banni Satria Andoko

State Polytechnics of Malang, Brawijaya University Malang


Abstract

A person with problems in the visual organs whose often called visually impaired person, either total or partial blindness has obstacles in doing motor activities, especially activities that require a person to move from a place to another such as walking have veering tendency. This problem can risk the safety of the visually impaired person when walking in public area. Thus the need for an application that can help people with special needs is felt important enough to support ease of activity for these people. This research proposes a mobile app based system to guide someone to walk straight using the motion sensors on the smartphone and audio-based guidance. The low-pass filter was applied to magnetometer and accelerometer data to reduce noise that resulted when the user walks. The results of this study can be used to train the visually impaired person to reduce the veering habits.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=136


80 ABS-155 Media, game and mobile technologies

Design of Public Transportation Navigation System on Android Wear Device
Komang Candra Brata (a*), Aryo Pinandito (b), Mahardeka Tri Ananta (a)

a) Mobile, Game, and Media Research Group,
Computer Science Faculty, Universitas Brawijaya
Malang, Indonesia
b) Information System Department,
Computer Science Faculty, Universitas Brawijaya
Malang, Indonesia


Abstract

Abstract?The application of information technology in the field of navigation in public transportation and the popularity of the Android operating system on mobile devices led to a variety of mobile application that is aiming to facilitate users in navigating with their desired mode of transportation. Google Maps application, which is provided by Android-based mobile devices, is currently able to provide information and navigation on city?s public transportation although its availability is still limited to certain areas and cities. Its limitation brings up mobile device navigation applications to be built by local developers to provide information and navigation in using public transport in such area. When people are navigating on public transport by utilizing mobile device applications, they may experience difficulties in operating their mobile devices or are not allowed to use their smartphone during the trip. Finding stops or interchanges to take or how long they have to stay on the line before they have to get off or switch to another line become more difficult. To overcome these problems, this research proposed a design and evaluate a new input interaction method that utilizes Android Wear-based wearable devices in implementing an intuitive local public transport navigation system. The proposed system is expected to step-by-step assist users action in public transport navigation by utilizing Android Wear-based wearable devices such as Android smartwatch and its embedded sensors. Through a user study, experimental results show that this simplification allows step-by-step directions for pedestrian navigation using smartwatches is feasible for future implementation but have different level of usability compared to a traditional mobile map application.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=155


81 ABS-178 Media, game and mobile technologies

Analysis of Web Content Delivery Effectiveness and Efficiency in Responsive Web Design Using Material Design Guidelines and User Centered Design
Aryo Pinandito (a*), Hanifah Muslimah Az-zahra (a) , Lutfi Fanani (b), Anggi Valeria Putri (a)

a) Mobile, Game, and Media Research Group,
Information System Department,
Computer Science Faculty, Universitas Brawijaya
Malang, Indonesia

b) Mobile, Game, and Media Research Group,
Computer Science Department,
Computer Science Faculty, Universitas Brawijaya
Malang, Indonesia


Abstract

Nowadays, websites are not only being accessed by computers with a large screen. They were mostly accessed via mobile devices such as tablets and smartphones that relatively have smaller screen size. Almost every website has their own style and visual appearance in terms of content and information delivery. Even though they were visually looked different, they were mostly designed and developed by using one of one column, two columns, or three columns layout. The way to properly deliver content and information on a smaller screen size becomes a challenging matter. Responsive Web Design approach allows a single web page to be differently visualized based on the screen size of the accessing device. Such layout changing may affect the amount of information to be displayed on the screen, hence affecting the effectiveness and efficiency of information delivery in a web page. This study compares the effectiveness and efficiency of a web page that is being displayed on a computer screen, tablet, and smartphones. How the implementation of Materials Design Guidelines and User-Centered Design approaches in the design process may affect the effectiveness and efficiency of content and information delivery of a web page on smaller screens were also evaluated and presented. This research shows that User Centered Design and Material Design Guidelines improve the effectiveness and efficiency of content delivery on both tablet and smartphone. Materials Design Guidelines provide better usability improvement than User Centered Design approach for smartphone in terms of content delivery efficiency for web pages designed in three columns layout.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=178


82 ABS-196 Media, game and mobile technologies

Numerical Analysis and Design of Inverted L Antenna for UHF TV receiver Application
Chandrasena Setiadi(a*), Erfan Rohadi(b), Moechammad Sarosa(c), Isa Mahfudi(d*)

(a)(d)Graduate School of Electronics Engineering Department, The State Polytechnic of Malang, Jl. Soekarno Hatta No. 9, Malang 65141, Indonesia
*chandrasenasetiadi[at]polinema.ac.id, isa_mahfudi[at]polinema.ac.id
(b)(c) Lecture of Electronics Engineering Department
The State Polytechnic of Malang, Jl. Soekarno Hatta No. 9, Malang 65141, Indonesia


Abstract

Generally, antennas functioned as a television receiver has big dimensions with quiet difficulty installation process. In this research, the proposed Inverted L antenna has been numerically analyzed to design the television antenna receiver at 639 MHz. The characteristic of proposed IL antenna has been investigated by adjusting the height of the IL antenna (h), the size of the conducting plane (px ? py) and the horizontal element length (L and L1) is optimized so that the input impedance matches the designed frequency. The inverted L antenna falls from the coaxial cable with the inner and outer radius respectively 0.255 mm and 1.095 mm. In the calculation, the height of the antenna (h) is adjusted from 15.6 mm up to 27.6 mm, the size of conducting plane (px ? py) was 80 mm x 200 mm, 100mm x 250mm, 200mm x 350mm and the length of horizontal elements are optimized so that the input impedance at 50 Ohm. In the calculation, the electromagnetic simulator WIPL-D Simulator is used. When the size of conducting plane 100 x 250 mm, the height of the IL antenna 15.6 mm, the length of horizontal elements L1 = 85 mm, L = 113.2 mm, the return loss bandwidth characteristic (-10dB) of 1.6% and the directivity gains of 3.58 dBi have been achieved at the resonant frequency of 639MHz. The proposed IL antenna promises for the television antenna system. The future work will be addressed to enhance the bandwidth of the proposed antenna.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=196


83 ABS-25 Natural language processing

Road Traffic Topic Modeling on Twitter using Latent Dirichlet Allocation
Ahmad Fathan Hidayatullah (a), Muhammad Rifqi Maarif (b)

a) Department of Informatics, Universitas Islam Indonesia
Jl. Kaliurang KM 14,5 Sleman, Yogyakarta, Indonesia

b) Department of Informatics Management, STMIK Jend. Achmad Yani Yogyakarta
Jl. Ringroad Barat, Banyuraden, Gamping, Yogyakarta, Indonesia


Abstract

Twitter has been widely used by some institutions in Indonesia to spread important information to the public. TMC (Traffic Management Center) as the one of the part in the Indonesian National Police has utilized Twitter as the medium to share about traffic information to the society. This research aims to create the topic model regarding traffic information on Indonesian Twitter messages. The data used in this research are retrieved from the official Twitter account of the Traffic Management Center in Java. LDA (Latent Dirichlet Allocation) is the method to build the topic model from the dataset. The topic model obtained will represent what kind of topics which posted by TMC in each region in Java. Thus, the result of this experiments could illustrate the problems and important information that happened in Java.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=25


84 ABS-53 Natural language processing

Deep Belief Network Optimization in Speech Recognition
Murman Dwi Prasetio (a*), Tomohiro Hayashida (b), Ichiro Nishizaki (c), Shinya Sekizaki (d)

a). Graduate School of Engineering, Hiroshima University
Higashi-Hiroshima, 739-8527, Japan
d161685[at]hiroshima-u.ac.jp
b). Graduate School of Engineering, Hiroshima University
Higashi-Hiroshima, 739-8527, Japan
hayashida[at]hiroshima-u.ac.jp
c). Graduate School of Engineering, Hiroshima University
Higashi-Hiroshima, 739-8527, Japan
nisizaki[at]hiroshima-u.ac.jp
d). Graduate School of Engineering, Hiroshima University
Higashi-Hiroshima, 739-8527, Japan
sekizaki[at]hiroshima-u.ac.jp


Abstract

The paradigm theory of neural network (NN) is expressing from biological of human brain function system. In this study, we deal with the deep belief network (DBN) model using the speech as natural language processing for training the system. The characteristic of deep belief network is similar with neural network. Whereas, the performance of neural network depends on the structure itself and it is suitable to select the model and size of the network for the data to handle. In occasion, deep belief network has the advantage in speech recognition it can be generated the feature learning with a subsequent stage of supervised learning, in which, the training of each layer weights were first initializing by unsupervised way and then fine-tuned with the labeled data. Deep belief network (DBN) have many nonlinear hidden layers to produce posterior of probabilities that takes several frames of coefficient as input. This paper shows the good performance improvement of modularity in DBN as combination technique of computation.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=53


85 ABS-127 Natural language processing

Classification Method Comparison on Indonesian Social Media Sentiment Analysis
Tirana Noor Fatyanosa; Fitra A. Bachtiar

Faculty of Computer Science, Universitas Brawijaya


Abstract

Sentiment analysis from social media has turned out to be essential since individuals are normally genuine with their sentiment on giving their perspective. However, in turning social media into a sentiment analysis possess challenges such as comments are usually ambiguous, language barrier problem, slang words, redundant comment, and sentiment classification. This study attempted to distinguish the issues of sentiment classification from Indonesian social media on Jakarta governor election. Several steps are taken to overcome those problems that include preprocessing. The preprocessing strategy used are removing the unrelated tweet, removing URL, deleting duplicate lines, deleting similar lines, removing the unrelated word, removing hashtag, removing Twitter username, removing number in the comment, removing punctuation, checking slang words, and converting the slang word into appropriate word. The preprocessed sentiment is then classified into positive, negative, and neutral. The classification method used in this study are Summation method, Average on Tweet, and Average on Tweet with the threshold on objective score, Weighted Average, and Naive Bayes method. The experimental result shows that the best classification method is Average on Tweet with the threshold on the objective score that yields 58.6% precision.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=127


86 ABS-153 Natural language processing

Time Series Neural Network Model for Part-of-Speech Tagging Indonesian Language
Theo Tanadi

School of Electrical Engineering and Informatics
Institut Teknologi Bandung


Abstract

Part-of-speech tagging (POS tagging) is an important part in natural language processing. Many methods have been used to do this task, including neural network. This paper models a neural network that attempts to do POS tagging. A time series neural network is modelled to solve the problems that a basic neural network faces when attempting to do POS tagging. In order to enable the neural network to have text data input, the text data will get clustered first using Brown Clustering, resulting a binary dictionary that the neural network can use. To further the accuracy of the neural network, other features such as the POS tag, suffix, and affix of previous words would also be fed to the neural network.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=153


87 ABS-28 Robotic systems

Performance Analysis of A* Algorithm to Determine Shortest Path of Fire Fighting Robot
Akhmad Alfan Hidayatullah, Anik Nur Handani, Muhammad Jauharul Fuady

State University of Malang


Abstract

A* Algorithm is one of Best First Search Algorithm, that combines Uniform Cost Search and Greedy Best-First Search Algorithm. This paper will discuss performance analysis of A* Algorithm in the case of KRPAI (Indonesian Fire Fighting Robot Contest). Time complexity and space complexity is the criteria that we use in this paper. Time complexity is the time needed by algorithm to process the command. Whereas, space complexity is the memory usage by algorithm to process the command. Based on the experiments, A* Algorithm need an average of 4.270,72 and 14.192 bytes of memory to find a solution. We expect a lot of future search will improve the mapping capability of firefighting robot. Therefore, someday firefighting robot can be implemented in real life, and will reduce the fire disaster victims.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=28


88 ABS-151 Robotic systems

A Control Scheme for Typist Robot using Artificial Neural Network
Wahyu K. Dewanto, Syamsiar Kautsar*, Khafidurrahman Agustianto

Information Technology Departement
Politeknik Negeri Jember
Jalan Mastrip, Kab. Jember
*kautsar.sam[at]gmail.com


Abstract

In 2005, UNICEF estimated the number of children with disabilities under age 18 at 150 million. Indonesia had 11 million workers with disabilities. It was less than 50% of the total number of disabilities person (data in 2010). Various efforts have been made to help disabilities person to be able to work normally. There are a lot of researches of prosthetic limbs, artificial hands, and motorized wheelchairs. In this paper, a typist robot was built. It was designed for people with physical hand disability. It helped disability people to operate computer normally. The typist robot consists of 2 arm robots. Each arm has 3 degree of freedom (DOF). An IMUsensor is mounted on the user?s foot. It?s used to measure the user?s foot movement. A mini USB keyboard is used as the working object of the robot. Arfiticial Neural Network (ANN) was used to convert the user?s foot movement into arm robot movement. The implementation of ANN for typist robot controlling had a success rate of 82%.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=151


89 ABS-82 Smart city

A Systematic Review of Conceptual Frameworks of Smart City Initiative
Ruci Meiyanti (a*), Dana Indra Sensuse (b)

a), b) Faculty of Computer Science, University of Indonesia
Jl. Margonda Raya, Beji, Pondok Cina, Kota Depok, Jawa Barat 16424
* ruci.meiyanti[at]ui.ac.id


Abstract

The smart city initiative that the first potential thing is the early stage of the smart city projects development. It is revealed by the conceptual frameworks. The conceptual frameworks of smart city initiative are the useful references to the developing project of the smart cities. There are researches about the variant conceptual frameworks of the smart city initiative. The purpose of this study is to increase the insight of the conceptual frameworks of smart city initiative. The method that is used in this study is the systematic review of Software Engineering and added with the Systematic literature review of Information Systems research. To do the systematic review, we collect the diversity of conceptual frameworks of smart city initiative from journal and conference publications indexed by Scopus in the last five years with related topics. The total distribution of smart cities topic was 162 literature. Then they were selected by the smart city frameworks became 30 literature, and finally selected by the conceptual frameworks of smart city initiative became 5 literature. The result of this research is fifteen elements that usually are used to construct in the conceptual frameworks of smart city initiative. They are expected to be a choice to make the appropriate conceptual frameworks of smart city initiative.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=82


90 ABS-42 Smart sensor networks

Implementation of Semantic System In The Smart Home Lights Device Based On Agent
Agung Prasetio (a*), Sabriansyah Rizqika Akbar (a), Bayu Priyambadha (a)

a) Faculty of Computer Science, Brawijaya University
Jalan Veteran, Malang 65145, Indonesia
*agungofficial[at]gmail.com


Abstract

The IoT device provides the potential for obtaining data contextually, by observing and measuring the physical events occurring in the environment. Smart Home has great interest by IoT device developers. Using a semantic approach to data processing on IoT devices provides the ability to define, modify, and interpret better data. In this research uses Nodemcu ESP8266 as a microcontroller on sensor node and actuator node. Raspberry Pi3 is used as an agent that serves as a router and data collection center. Both devices already support wifi connectivity so as to facilitate the communication process. The sensor node is responsible for collecting data derived from the sensors process of the sensor module, the data obtained is sent to the agent to be initialized and inserted into the ontology. Agent performs the reasoning process to change existing data on the ontology according to the rule created and produce more representative and contextual data. The Actuator node performs a query against the current status of the lamp on the ontology as a reference for the action of switching on or off the lamp. Implementation in the form of lights on and off control system in smart home with semantic approach.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=42


91 ABS-44 Smart sensor networks

Message Queue Telemetry Transport Protocols Implementation for Wireless Sensor Networks Communication - A Performance Review
Sabriansyah Rizqika Akbar (a*), Kasyful Amron (a), Harry Mulya(a), Sofi Hanifah (a)

a) Faculty of Computer Science, University of Brawijaya
Jalan Veteran No.8 , Malang 65145, Indonesia
*sabrian[at]ub.ac.id


Abstract

This Paper presents the result of Message Queue Telemetry Transport (MQTT) Protocols implementation in a Wireless Sensor Network (WSN). Arduino and NRF were act as WSN devices. Nodes were equipped with sensor modules and NRF modules as communication interface. While the gateway that implemented on top of Arduino Pro-mini, was completed with NRF24L01 module for the WSN communication interface, and ESP8266 module as the IP based Wi-Fi interface to the MQTT Broker. Sensor node and gateway communicate to each other by using MQTT-SN protocols with the QoS 0, QoS 1, and QoS 2 publish type as the communication performance parameters of the WSN. Packet loss measurement was done in QoS0, data duplication and re-tries in the QoS1 & QoS2, and round-trip time delay measurement in the QoS1 and QoS2. This project also considers gateway discovery process, as part of communication scheme.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=44


92 ABS-30 Software engineering

Development of an Adaptive interval value fuzzy number based on MCGDM model by Hybrid AHP and TOPSIS methods
Yeni Kustiyahningsih

Universitas Trunojoyo


Abstract

In this research, a new adaptive interval value fuzzy multiple criteria group decision-making model is developed by integrating Analytic Hierarchy Process (AHP) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The purpose of this research is to provide more accurate modeling, and performance rating better. Interval Value Fuzzy has a more effective representation, high flexibility, and more efficient computing. The modeling and representation of linguistic expressions in the form of interval fuzzy have an accuracy better than Fuzzy. The AIVFAHP approach is capable of elaborating complex multi-criteria problems into a hierarchical structure, considering the value of logical consistency in judgment and determining the optimal weight in multi-criteria decision-making. In determining the best alternative, the AIVFTOPSIS approach is well-suited for being able to make multi-decisions based on the concept that the chosen alternative has the shortest distance from the ideal solution and furthest from the ideal negative solution. The method is implemented in a Strengths, Weaknesses, Opportunities, and Threats (SWOT) based strategy selection problem for measurement e-learning.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=30


93 ABS-96 Software engineering

Implementation of mobile-based monitoring sales system in Semi Tani Shop
Hendrik Setyo Utomo, Rabbini Sayyidati, Oky Rahmanto

Politeknik Negeri Tanah Laut


Abstract

Sales transaction is an activity undertaken to sell goods, for example: agricultural medicines, solid fertilizer, liquid fertilizer, and others in cash or credit. Semi Tani Shop is one of agricultural store located in Tanah Laut. Semi Tani Store has a sales transaction management, inventory stock, and profit calculation done conventionally. The way that is done is using paper in the calculation of income statement and recording stock of goods. However, there is no bookkeeping sales transaction in any form and it makes hard to monitor the transaction history. This study aims to implement sales information system of agricultural store sales based on Mobile and sales data stored in local computer and store in cloud. The sales information system focuses on monitoring reports that can be accessed anywhere with the condition connected to the internet and have POS cloud system i.e OSPOS (Open Source Point of Sales). The report is a Goods Stock report, Transaction Report and income statement. The method for this research is using Prototype. Implementation results help Semi Tani Shop owners in sales management using OSPOS, as well as monitoring of stocks, stock notes in specified quantities (thinning), transaction reports, and income statement with mobile android.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=96


94 ABS-110 Software engineering

Pioneering the Automation of Internal Quality Assurance System of Higher Education (IQAS-HE) Using DevOps Approach
Acep Taryana; Iwan Setiawan; Eko Bayu; Ari Fadli

Unsoed


Abstract

It was important to develop a software for managing the Internal Quality Assurance System for Higher Education (IQAS-HE). In addition, automation was a hot issue to develop the software that had capability to present some indicators of the quality assurance into a database. Possibly, some indicators were formulated when IQAS-HE was executed, and the indicators came be some new indicators for database. How to implement the indicators in the IQAS-HE that could be a part of database in the software for measurement the IQAS-HE automatically. Therefore, this research present the automation for developing a management tools required by higher education institutions in terms of quality improvement, using the DevOps approach. Furthermore, DevOps was an emerging paradigm for eliminate the split and barrier between developers and operating personnel that traditionally exist in many companies today. In this research, management tools aim to bridge the gap between stakeholders (customers) and developers to present software that had a good quality that could be a tool for improving the process of quality assurance of higher education continously. The research result show that a programming technique, which collaborated with in DevOps approach, was an important issue to obtain automation of IQHS-HE development.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=110


95 ABS-113 Software engineering

THE USE OF CURVE WIZARD METHOD FOR GENERATING ROUGH SURFACE IN CAD MODEL AND DATA TRANSFERING ANALYSIS IN CAE MODEL
1,2Kartini*, 2G.A. Sipayung, 2R.R.S. Wicaksono, 2E. Saputra, 2R. Ismail, 2J. Jamari, 2A.P. Bayuseno

1 Department of Informatics Engineering, Universitas Pembangunan Nasional "Veteran" Jawa Timur
Jl. Raya Rungkut Madya, Surabaya, 60294, Indonesia
2 Department of Mechanical Engineering, Diponegoro University
Jl. Prof. Sudharto Kampus UNDIP Tembalang, Semarang, 50275, Indonesia


Abstract

The contacting surfaces in the biomedical implants, e.g. artificial hip and knee impants, have a rough surface in the micro scale. The computational modelling of the contacting surface between implants with respect to the rough surfaces require are conducted using finite element modelling in CAE software. The rough surface of the contacting components can be modelled as a homogeneous rough surface which derived from equation in Matlab. Then the rough surface is exported to CAD and CAE model. A geometrical deviation can be appear during data transfer in the CAD and CAE model. This paper aims to study the geometrical analysis of the data transfer from mathematical models in Matlab to CAD models in SolidWorks and CAE models in ABAQUS software. In the CAD model, especially in SolidWorks, the rough surface generation process, which is imported from other file format, can be performed in 3 modes: automatic creation method, guided creation method and curve wizard method. The geometrical analysis is conducted on a sinusoidal rough surface model, especially at the peak of height, the valley depth, and the rough surface wavelength with respects to the three types of surface generation method. The result show that homogeneous rough surface successfully transferred to CAD model and CAE model using curve wizard method with the highest geometrical accuracy. The deviation in this method is less than 1 percent on average on 4 locations on the sinusoidal rough surface model. In the automatic creation method, and the guided creation method, the resulted rough surface is less acurate which is indicated by the higher geomterical deviation.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=113


96 ABS-132 Software engineering

A COMPARATIVE STUDY OF FILE TYPE SELECTION DURING DATA TRANSFER OF HOMOGEN ROUGH SURFACE FROM CAD MODEL TO CAE MODEL
G.A. Sipayung, R. Ismail, Kartini, E. Saputra, J. Jamari, A.P. Bayuseno

1 Department of Mechanical Engineering, Diponegoro University
Jl. Prof. Sudharto Kampus UNDIP Tembalang, Semarang, 50275, Indonesia
2 Department of Informatics Engineering, Universitas Pembangunan Nasional Veteran Jawa Timur
Jl. Raya Rungkut Madya, Surabaya, 60294, Indonesia


Abstract

Some types of rough surfaces are usually modeled first in CAD software. Rough surfaces that have been made will be processed for analysis through various methods, one of which is CAE software. To continue to CAE software, various file types can be used. This study discusses the selection of file extensions to the analysis process itself. Rough surfaces are imported first with various file type types, ie solidpart file (* sldprt), iges (* igs) file, step file (* stp), parasolid file (* x.t), and acis file (* sat). In this study, CAE software used is abaqus, while the model used is a homogeneous rough surface model. The five files of each file type are then inputted into CAE software and performed static contacts. The result of the static contact of each file is then compared with the three review parameters, namely CAD file size, ODB file size, and CAE file size. The result shows that file type differences do not make a geometry difference. Furthermore, the contacted and obtained surface is the parasolid file (* x.t) superior to the other four file types, obtained from the file size data of each smaller parameter. File size on parasolid file type is 57% smaller than file size in other file type, where iges file is file type with largest file size. So in the next study, the research team suggested the use of file type parasolid type.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=132


97 ABS-176 Software engineering

Developing Food Sensory Test System with Preference Test (Hedonic and Hedonic Quality)
Prawidya Destarianto, Hendra Yufit Riskiawan, Khafidurrohman Agustianto, Syamsiar Kautsar

Politeknik Negeri Jember


Abstract

Bread is a food source of carbohydrates that are often consumed by the community. Various types of bread was produced to meet consumers curiosity, one of which is wheat bread. Manufacturers must be able to produce quality wheat bread and liked by consumers. Increasing the quality of bread will certainly have an impact on sales to be generated. One of the efforts in improving the quality of wheat bread is by doing Hedonic test and Hedonic Quality test. This study aims to develop a system capable of providing an assessment of wheat bread. This study develop machine learning system with supervised learning algorithm, then using the results of the initial Organoleptic test as Knowledge Based (KB). This test involved detection, recognition, discrimination, scaling and ability to express likes or dislikes (hedonic quality), using K-Means Clustering with expert judgment. Hedonic quality is used as a variable for assessing wheat bread products with 4 variables, which include flavor, taste, appearance, and texture. While the hedonic test using two classes: likes or dislikes. This KB used as Naive Bayes Classifier algorithm initial knowledge, The test results using 10 fold shown average accuracy 93.34%, while the final goal of the development of this system will create a system capable of providing an assessment of a wheat bread product.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=176


98 ABS-9 Sustainable Technology

Optimizing Single Low-End LAMP Server Using NGINX Reverse Proxy Caching
Mahendra Data [1], Muhammad Luthfi [2], Widhi Yahya [3]

[1][2][3] Computer Science Faculty,
Brawijaya University,
8 Veteran Road, Malang 65145, Indonesia
[1] mahendra.data[at]ub.ac.id
[2] muh.luthfi9[at]ymail.com
[3] widhi.yahya[at]ub.ac.id


Abstract

This research aims to optimize single low-end Linux Apache MySQL PHP (LAMP) server. This kind of server usually used by newly born Indonesian startup that usually has a limited budget. We choose NGINX as reverse proxy caching for optimizing this single low-end LAMP server. The reason why we do not use NGINX as the web server, despite the fact that NGINX has better performance by many reviews, is because some popular web applications and framework only support Apache. This dependency force the startup owner to use Apache rather than NGINX as the web server. To evaluate the performance of this infrastructure, we use WordPress as the web content and then test it using two test. First, we test the server using 102 concurrent connections, then we increase it to 408 concurrent connections. Those scenarios have tested on a Virtual Private Server (VPS) hosted in DigitalOcean which has single 2.00 GHz CPU and 512 of RAM that cost only USD 5 per month. Experiment result shows that the proposed architecture can reduce page loading time up to 96%, saving CPU load up to 99% and saving memory usage up to 28%. We conclude that using NGINX as a reverse proxy on LAMP server is a good alternative solution to optimize the performance of the web server.

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=9


99 ABS-48 Sustainable Technology

Persuasiveness Web 2.0 in Behavioral Intention: A Conceptual Model
Yohana Dewi Lulu Widyasari(1,2),Lukito Edi Nugroho(1), Adhistya Erna Permanasari (1)

1.Electrical Engineering and Information Technology
Universitas Gadjah Mada, Yogyakarta
2.Information System,
Politeknik Caltex Riau, Pekanbaru


Abstract

The use of web 2.0 technology as a work tool or communication tool in individual activities grow significant and has begun to be also used in the organization. The Web 2.0 technology is not just an information technology system but also a social technology, which is still a debate of factors that can influence it. Aspects of social interaction owned by this technology impact on the behavior and attitude of the users. This research proposes a conceptual model intervention of persuasive technology that explains what the factor affect the users? behavioral intention to use technology in an organization. This study reviews of web 2.0 technology by using Theory of Planned Behavior and Behavior Change Support Systems to analysis its persuasiveness. In the present study, we tested a persuasive systems design model that had a significant impact on perceived persuasiveness and system usage. More specifically, the proposed model extends Theory of Planned Behavior to include seven constructs namely, Perceived Ease of Use, Perceived Usefulness, Attitude Toward The Behavior, Social Influence, Subjective Norm, Self-Efficacy, Perceived Behavioral Control. This model provides valuable insights into the factors that influence the acceptance or resistance of web 2.0 as persuasive technology by intended users and offers opportunities for future research in understanding the acceptance of persuasive technology

PermaLink: http://siet2017.interconf.org/pages/abstract.php?id=48


SIET 2017 - Submission Management System

Powered By Konfrenzi 1.828R © 2017 All Rights Reserved