Jurnal Pilar Nusa Mandiri https://ejournal.nusamandiri.ac.id/index.php/pilar <p>The PILAR Nusa Mandiri Journal is a formation of the Information Systems study program, which was originally a medium for accommodating scientific writings of STMIK Nusa Mandiri Jakarta Information Systems lecturers. Along with the times, this journal has become a National journal that has P-ISSN: 1978-1946 and E-ISSN: 2527-6514. PILAR Nusa Mandiri has become a Rank 3 Accredited Journal&nbsp;and is trying to become a higher accredited journal. PILAR Journal Nusa Mandiri is published 2 times in 1 year, namely in March and September. This journal is&nbsp;<span class="tlid-translation translation"><span title="">Rank 3 Accreditation Certificate (S3), Accreditation is valid for 5 years. Starting from Vol. 12, No. 1 the Year 2016 to Vol. 16, No. 2 the Year 2020.&nbsp;Journal of PILAR Nusa Mandiri, re-accreditation remains at Rank 3 (SINTA 3), starting Vol. 15 No. 2 of 2019 based on the Decree of the Minister of Research and Technology / National Research and Innovation Agency Number 85 / M / KPT / 2020, April 1, 2020</span></span></p> PPPM Nusa Mandiri en-US Jurnal Pilar Nusa Mandiri 1978-1946 <div class="page"> <p>The Authors submitting a manuscript do so on the understanding that if accepted for publication, copyright of the article shall be assigned to the PILAR Nusa Mandiri journal as the publisher of the journal, and the author also holds the copyright without restriction.</p> <p>Copyright encompasses exclusive rights to reproduce and deliver the article in all form and media, including reprints, photographs, microfilms, and any other similar reproductions, as well as translations. The reproduction of any part of this journal, its storage in databases, and its transmission by any form or media, such as electronic, electrostatic and mechanical copies, photocopies, recordings, magnetic media, etc. , are allowed with written permission from the PILAR Nusa Mandiri journal.</p> <p>PILAR Nusa Mandiri journal, the Editors and the Advisory International Editorial Board make every effort to ensure that no wrong or misleading data, opinions, or statements be published in the journal. In any way, the contents of the articles and advertisements published in the PILAR Nusa Mandiri journal are the sole and exclusive responsibility of their respective authors and advertisers.</p> </div> MEASUREMENT OF VALIDITY AND RELIABILITY OF CUSTOMER SATISFACTION QUESTIONER in E-BOARDING APPICATIONS https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1069 <p>PT. Kereta Api Indonesia (Persero) is the largest company in Indonesia which is engaged in railroad transportation. One of the technological innovations carried out by PT KAI (Persero) is to make the KAI Access application with the latest feature is e-boarding. Through the e-boarding feature provided by PT KAI, KAI seeks to provide excellent service to customers. Not many customers use the e-boarding application as a gap between PT KAI's vision and train customers. The gap that exists between the vision and customer acceptance of information technology needs to be measured to determine the extent to which the implementation of the information technology is accepted by the customer. One method used to measure customer satisfaction is to use the use questionnaire. Before measuring customer satisfaction using the Use questionnaire, each item of questions must be tested for validity and reliability. The purpose of measuring the validity and reliability of this research is to ensure the accuracy and reliability of each question on the questionnaire as a test tool. The validity test results used by using the construct validity technique are valid. The reliability test results using the split-half reliability technique have a value of 0.9896 or equivalent to the very high-reliability criteria.</p> Sisilia Thya Safitri Dwi Mustika Kusumawardani Citra Wiguna Didi Supriyadi Intan Yulita ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 1 6 10.33480/pilar.v16i1.1069 ANALYSIS OF KARAWANG ONLINE SALES CUSTOMER SATISFACTION USING CUSTOMER SATISFACTION INDEX (CSI) METHOD https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1111 <p>Karawang is one of the industrial cities. Most industry players look at Karawang as a strategic city to run a business. Many products have been produced from Karawang. However, there are lack in promoting, marketing the product and expanding the marketing area. The analysis of consumer satisfaction in Karawang is to determine the satisfaction of Karawang consumers to the prospects of promising online sales. Service attributes can be included in increasing online sales at Karawang using the Customer Satisfaction Index (CSI) method. The result of the Customer Satisfaction Index (CSI) is 78.43% which means that overall consumers who live in Karawang and have been shopped online are satisfied with the development of online shopping. This research was conducted in Karawang. The data used are primary data and secondary data. The sampling method is a non-probability sampling method, while the non-probability sampling method used sampling purposes.</p> Hannie Hannie Ultach Enri Yuyun Umaidah ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 7 12 10.33480/pilar.v16i1.1111 DECISION SUPPORT SYSTEM USING AHP METHOD FOR TEACHER PERFORMANCE ASSESSMENT https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1031 <p>One effort to measure the level of quality in schools is by measuring the performance aspects of teachers as professional educators who teach at the school. The teacher performance aspect is measured as one of the promotion requirements for a higher position or as a recommendation condition in order to take part in teacher certification activities. In order for teacher performance appraisal to be carried out objectively, a method is needed to assist in the teacher performance appraisal process. AHP method can be used to assist in decision making. This is because the AHP method is a model for structured and comprehensive decision making. From the calculation using the AHP method, it was found that the first priority was obtained by Indra with a weight of 0, 7317 or 73.17%, the second priority was obtained by Reni with a weight value of 0.2279 or 22.79% and the lowest priority was obtained by Supriyatna with a weight value 0.0604 or 6.04%.</p> Rahmawati Rahmawati Dewi Ayu Nur Wulandari ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-02 2020-03-02 16 1 13 18 10.33480/pilar.v16i1.1031 IMPLEMENTATION OF GAIN RATIO AND K-NEAREST NEIGHBOR FOR CLASSIFICATION OF STUDENT PERFORMANCE https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/813 <p>Predicting student performance is very useful in analyzing weak students and providing support to students who face difficulties. However, the work done by educators has not been effective enough in identifying factors that affect student performance. The main predictor factor is an informative student academic score, but that alone is not good enough in predicting student performance. Educators utilize Educational Data Mining (EDM) to predict student performance. KK-Nearest Neighbor is often used in classifying student performance because of its simplicity, but the K-Nearest Neighbor has a weakness in terms of the high dimensional features. To overcome these weaknesses, a Gain Ratio is used to reduce the high dimension of features. The experiment has been carried out 10 times with the value of k is 1 to 10 using the student performance dataset. The results of these experiments are obtained an average accuracy of 74.068 with the K-Nearest Neighbor and obtained an average accuracy of 75.105 with the Gain Ratio and K-Nearest Neighbor. The experimental results show that Gain Ratio is able to reduce the high dimensions of features that are a weakness of K-Nearest Neighbor, so the implementation of Gain Ratio and K-Nearest Neighbor can increase the accuracy of the classification of student performance compared to using the K-Nearest Neighbor alone.</p> Tyas Setiyorini Rizky Tri Asmono ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-02 2020-03-02 16 1 19 24 10.33480/pilar.v16i1.813 ANALYSIS OF PLANT FRAGARIA XANANASSA DISEASE DIAGNOSES USING PRODUCTION RULES BASE ON EXPERT SYSTEM https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1174 <p>Errors that occur in solving problems in strawberry plants (Fragaria Xananassa) such as the presence of leaf patches, fruit rot, perforated leaves, and insect pests can be the cause of not maximum in harvest time. The farmers and the general public who planted strawberry (Fragaria Xananassa) need to know the proper treatment of diseases and pests so that future yields as expected. Therefore, it takes an application as a solution in the delivery of information related to the problems that are often encountered in strawberry plants (Fragaria Xananassa). Methods of production rules can be used to diagnose the disease strawberry (Fragaria Xananassa) based on signs or symptoms that occur in the parts of plants and strawberry, the results of diagnosis using this method are the same as we do Consultation on experts. &nbsp;The purpose of this study was to determine the early diagnosis of disease in strawberry plants (Fragaria Xananassa) based on signs or symptoms that occur in the plant and fruit parts. The results of the analysis of this study showed that the validation of disease and symptom data in strawberry plants (Fragaria Xananassa) reached 99%, meaning that between the data of symptoms and disease understudy the accuracy was guaranteed with the experts.</p> Basiroh Basiroh Wiji Lestari ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-02 2020-03-02 16 1 25 32 10.33480/pilar.v16i1.1174 IMAGE BACKGROUND PROCESSING FOR COMPARING ACCURACY VALUES OF OCR PERFORMANCE https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1076 <p><em>Optical Character Recognition (OCR) is an application used to process digital text images into text. Many documents that have a background in the form of images in the visual context of the background image increase the security of documents that state authenticity, but the background image causes difficulties with OCR performance because it makes it difficult for OCR to recognize characters overwritten by background images. By removing background images can maximize OCR performance compared to document images that are still background. Using the thresholding method to eliminate background images and look for recall values, precision, and character recognition rates to determine the performance value of OCR that is used as the object of research. From eliminating the background image with thresholding, an increase in performance on the three types of OCR is used as the object of research.</em></p> Desiana Nur Kholifah Hendri Mahmud Nawawi Indra Jiwana Thira ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 33 38 10.33480/pilar.v16i1.1076 SELECTION OF EXTRACURRICULAR ACTIVITIES IN SMK INSAN AQILAH 4 JAKARTA USING PROFILE MATCHING METHOD https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/913 <p>This research is based on the observations and experiences of researchers, because the many extracurricular activities in schools make it difficult for students to determine extracurricular activities that can develop their personality, talents, and abilities outside the academic field, therefore the system is created to select extracurricular activities. The aim is to help students deal with extracurricular selection problems. The application of the Profile Matching method in the decision support system for the selection of extracurricular activities is expected to help provide recommendations for extracurricular activities to overcome the problem of selecting extracurricular activities and can facilitate students in selecting extracurricular activities.</p> Wahyudin Wahyudin Andi Saryoko Abdul Aziz Lia Nurmalia ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 39 44 10.33480/pilar.v16i1.913 DECISION SUPPORT SYSTEM IN DETERMINING THE BEST JUDO ATHLETE USING AHP METHOD https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/919 <p>To determine the best Judo athlete, Rajawali Judo Club of Battalion Arhanud 1 Divif 1 Kostrad has several obstacles such as making a decision in determining the best Judo athlete by the Coach and the Achievement Development which only based on experience which is estimated without the existence of any system. This results in subjectivity and the absence of a strong basis for competent objective decision making which then triggers gaps between athletes. The absence of specific criteria creates that result in not targeting the selection of the best Judo athletes. For this reason, a method, in this case, is needed to choose the AHP (Analytical Hierarchy Process) method and a number of criteria as indicators in determining the best Judo athlete. While the referenced criteria are Self-Dropping Technique (Ukemi), Slamming Technique (Nage-waza), Lockdown or Lying Technique (Katame-waza), Discipline and Achievement. The purpose of this study is expected to produce statistical data as an evaluation material for the training team to minimize or even eliminate the gap between fellow Judo athletes at the Rajawali Judo Club of Battalion Arhanud 1 Divif 1 Kostrad. The result of this study is based on Analytical Hierarchy Process calculations, obtained the most important priority criteria in determining the best Judo athlete in which the Achievement criteria with value &nbsp;0.325 or equivalent to 32%, then followed by Disciplinary criteria with value 0.227 or equivalent to 23%, Slamming Technique criteria (Nage-waza) with value 0.211 or equivalent to 21%, Lockdown/Laying Technique criteria (Katame-waza) with value 0.125 or equivalent to 12% and in the last rank the Self-Dropping Technique criteria (Ukemi) with value 0.112 or equivalent to 11%.</p> Dinar Ajeng Kristiyanti Garth Wishnuwardhana Pangemanan ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 45 52 10.33480/pilar.v16i1.919 HYBRID OPTIMIZATION METHOD BASED ON GENETIC ALGORITHM FOR GRADUATES STUDENTS https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1180 <p>Graduation is a target that must be achieved by students, especially graduating on time will be very important. To determine students who graduate on time or cannot be determined before students reach the final semester and hold a trial, many students who fail to graduate on time cause delays and affect the quality assurance of a tertiary institution. The problem in this research is how to optimize student graduation in order to graduate on time. Therefore, to determine this decision, we conducted a graduation data trial using the SVM method with GA optimization. SVM with accurate learning skills and good generalizations in classifying non-linear data, but SVM is weak in terms of parameter optimization it requires optimization using GA. GA is a method that has evolved to produce a more optimal data. From the results of processing using SVM and GA, we get more optimal results with 86.57%. Then from these results can help students to graduate on time<em>.</em></p> Ridwansyah Ridwansyah Ganda Wijaya Jajang Jaya Purnama ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 53 58 10.33480/pilar.v16i1.1180 COMPARISON OF MACHINE LEARNING CLASSIFICATION ALGORITHM ON HOTEL REVIEW SENTIMENT ANALYSIS (CASE STUDY: LUMINOR HOTEL PECENONGAN) https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1131 <p>Analysis of hotel review sentiment is very helpful to be used as a benchmark or reference for making hotel business decisions today. However, all the review information obtained must be processed first by using an algorithm. The purpose of this study is to compare the Classification Algorithm of Machine Learning to obtain information that has a better level of accuracy in the analysis of hotel reviews. The algorithm that will be used is k-NN (k-Nearest Neighbor) and NB (Naive Bayes). After doing the calculation, the following accuracy level is obtained: k-NN of 60,50% with an AUC value of 0.632 and NB of 85,25% with an AUC value of 0.658. These results can be determined by the right algorithm to assist in making accurate decisions by business people in the analysis of hotel reviews using the NB Algorithm.</p> Jaja Miharja Jordy Lasmana Putra Nur Hadianto ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 59 64 10.33480/pilar.v16i1.1131 A DIJKSTRA ALGORITHM IMPLEMENTATION IN DETERMINING SHORTEST ROUTE TO MOSQUE IN RESIDENTIAL CITRA INDAH CITY https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1199 <p>The application of artificial intelligence (Artificial Intelligence) for problem-solving in the field of computer science has experienced rapid development from year to year as the development of artificial intelligence itself. Problems involving searching (searching) is one example of the use of artificial intelligence that is quite popular to solve various kinds of problems. In daily activities, the use of roads is always an unavoidable activity, so determining the shortest path from one point to another becomes a problem that is often encountered. This is also felt by residents who live in a large enough housing. Sometimes to be able to reach the destination they are often confused in deciding which way to go to get the shortest distance to the destination. Citra Indah City Housing is a residential area in the Jonggol District area, Bogor Regency, developed by the Ciputra group. Within the Vignolia Hill Cluster, there is a mosque located on the northwest corner of the Vignolia Hill cluster or at the western end of the AH.17 block. A large number of blocks raise problems regarding the shortest route that can be taken by residents to get to the mosque. So, the purpose of this research is to determine the shortest path taken by citizens to get to the mosque. The method used is to apply the Djikstra algorithm which is able to produce the shortest route for residents to get to the mosque.</p> Siti Lestari Lestari Ardiansyah Ardiansyah Angelina Puput Giovani Desy Dwijayanti ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-15 2020-03-15 16 1 65 70 10.33480/pilar.v16i1.1199 DIAGNOSIS OF CORONAVIRUS DISEASE 2019 (COVID-19) SURVEILLANCE USING C4.5 ALGORITHM https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1293 <p>Coronavirus Disease 2019 (COVID-19) has become a pandemic in Indonesia as a non-natural disaster in the form of disease outbreaks which must be undertaken as a response. The Ministry of Health in the Republic of Indonesia published a guidebook for prevention and control of COVID-19 in its response efforts. This guideline is intended for health officials as a reference in preparing for COVID-19. This handbook contains early detection and response activities to identify conditions of PDP, ODP, OTG, or confirmed cases of COVID-19. The efforts made are adjusted to the world situation progress from COVID-19 which is monitored by the World Health Organization (WHO). From the results of documentation studies that have been carried out on the COVID-19 pandemic in Indonesia, there are several problems that must be resolved from the prevention of the disease outbreak COVID-19. Lack of knowledge and awareness of the general public in the prevention and control of COVID-19 is one of the factors increasing the spread of that virus in Indonesia. Furthermore, there are difficulties in carrying out surveillance, early detection, contact tracing, infection prevention or control, and risk communication or people empowerment. This is due to the lack of implementation and testing on artificial intelligence methods for COVID-19 diagnosis that can be used by the public. The purpose of this research is to make a diagnosis of surveillance classification which includes PDP, ODP, and OTG using the C4.5 algorithm. The results showed that the diagnosis of the COVID-19 surveillance category using the C4.5 algorithm was successfully modeled into a decision tree with PDP, ODP, and OTG classification. The testing process in a confusion matrix with 3 (three) classes produces an accuracy rate of 92.86% which is included in the excellent classification category.</p> Wildan Wiguna Dwiza Riana ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-16 2020-03-16 16 1 71 80 10.33480/pilar.v16i1.1293 SUCCESS ANALYSIS OF KITABISA MOBILE APPLICATION INFORMATION SYSTEM BY USING DELONE AND MCLEAN MODELS https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/918 <p>KitaBisa, the public can raise funds for a variety of things they do or donate for things they want to help. Starting from the desire to create a social movement, in 2013 Alfatih Timur (Timmy) made KitaBisa a forum for anyone who wanted to realize his social project. Measuring the success of the KitaC mobile application can use the DeLone and McLean model, it is known that variables that have a significant relationship include Service Quality against User Satisfaction with a t-statistic value of 2.893, System Quality for Users (Use) with a t-statistic value of 10,204, and User Satisfaction of the Net Benefit with a t-statistic value of 3,680. In accordance with the hypothesis testing that user satisfaction with the KitaBisa mobile application as a forum to donate online has been proven by influencing User Satisfaction with the Net Benefit as a whole or it means the user has felt the benefits of the KitaBisa mobile application.</p> Hafiz Noval Hasany Aditya Nurmalasari Nurmalasari Hendri Hendri ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-16 2020-03-16 16 1 81 88 10.33480/pilar.v16i1.918 ANALYSIS OF BUBBLE SORT AND INSERTION SORT ALGORITHM ON MEMORY EFFICIENCY USING DATA MINING APPROACH https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1165 <p>Sorting algorithm in the computational process makes it easy for users when the data sorting process because the data is sorted by the process quickly and automatically. In addition to speed in sorting data, memory efficiency must also be considered. In this research, a retesting of two sorting methods is conducted, namely the bubble sort method and the insertion sort method based on the comparison of two programming languages, Java with Visual Basic 2010 using the decision tree method. This research aims to find out which algorithm has lower memory consumption in the sorting process using Java or Visual Basic 2010. The results of the comparison show, in Visual Basic 2010. insertion sort algorithm which has the lowest average memory consumption of 4.3243KB for .vb extensions and 2.0145KB for .exe extensions. while the bubble sort method with a consumption amount of 4.4358KB for the .vb extension and 2.0352 for extension.exe. Furthermore, if you use the Java programming language. So the bubble sort method still consumes the highest average memory, which is 546,242KB for the .jar extension and 4,337KB for the .exe extension, whereas from the insertion sort method, which has a low average memory consumption of 543,578 KB for extension .jar, and 4,381KB for extension .exe<em>.</em></p> Iqbal Dzulfiqar Iskandar Imam Amirulloh Melisa Winda Pertiwi Mira Kusmira Agung Baitul Hikmah Deddy Supriadi ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 89 96 10.33480/pilar.v16i1.1165 PREDICTION OF GLUCOSE LEVEL IN DIABETICS WITH SUPPORT VECTOR REGRESSION https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1264 <p>One of the common diabetes factors that people hear is that they consume too much or often consume sweet foods or drinks so that blood sugar in the human body increases. The times and increasingly sophisticated technology make it easier for someone to be able to predict a disease such as diabetes with machine learning techniques. Therefore, from the existing problems, a machine learning technique will be made in predicting glucose levels in diabetics. The aim is to predict glucose levels in diabetics and find the best algorithm from several comparison algorithms. The results of the experiments carried out by the support vector regression algorithm have a lower mean squared error value of 28.9480 compared to other comparative algorithms and visualize the error classification seen that Instance no 47 has a prediction of the highest plasma glucose value of 189.2305.</p> Devi Wulandari Agus Subekti ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 97 102 10.33480/pilar.v16i1.1264 IMPLEMENTATION OF THE SAW METHOD AS A DECISION SUPPORT FOR GIVING FEASIBILITY OF KUR ON BANK MANDIRI DRAMAGA BOGOR https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1302 <p>Currently, the public's interest is very high to get KUR, but it makes it difficult for banks to determine who is eligible to receive the KUR and in the process of giving credit using the "LOS" system but this system is still quite a time consuming to analyze customer data and the process requires consideration and good analysis from the leader, due to the high number of problem loans. The SAW method used in this study. The SAW method is able to simplify and accelerate the results of credit lending recommendations. The calculation results obtained by debtors who are very worthy given credit as much as 1 debtor (4%), decent debtors with low risk as many as 16 debtors (70%), and worthy of being given with high risk as much as 6 debtors (26%) The purpose of this study to know the process and requirements for granting business credit at Bank Mandiri Dramaga Bogor.</p> Frieyadie Frieyadie Riki Setiyawan ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 103 110 10.33480/pilar.v16i1.1302 USTADZ ABDUL SOMAD LECTURE SENTIMENT ANALYSIS USING SUPPORT VECTOR MACHINE ALGORITHM COMPARISON OF COMPARATIVE FEATURES SELECTION https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/702 <p>Religious lectures are activities that are identical to the religious presentation, delivered verbally by a person who has religious knowledge and then delivered to the community with the aim of the knowledge delivered can be understood. Ustadz Abdul Somad was one of the preachers who had been known to various levels of society, but his lectures were not all acceptable to the people who liked or disliked those who came from various positive and negative comments on social media. To solve these problems, Sentiment Analysis was used by applying the Support Vector Machine Algorithm method. The purpose of this study is to compile using the selection of feature Particle Swarm Optimization and Information Gain. The results for Particle Swarm Optimization Selection Feature resulted in Accuracy of 80.57%, Precision of 85.45%, and Recall of 79.52%, Selection Feature Information Gain resulted in Accuracy of 79.78%, Precision of 78.47%, and Recall of 78, 43%, Based on the results of this study, it can be concluded that using the Particle Swarm Optimization selection feature is better at the level of accuracy when compared to using the Information Gain selection feature.</p> Dedi Aridarma Rifki Sadikin Bobby Suryo Prakoso Heru Sukma Utama ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 111 116 10.33480/pilar.v16i1.702 SENTIMENT ANALYSIS ON GOJEK AND GRAB USER REVIEWS USING SVM ALGORITHM BASED ON PARTICLE SWARM OPTIMIZATION https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1304 <p>Users of the Gojek and Grab application can provide reviews or comments about the application on Google Play. Reviews in the form of giving opinions about their satisfaction or dissatisfaction with the services provided. So with the many opinions provided, making people selective in choosing an online motorcycle taxi service provider. The application with the best review will be chosen by the community. In previous studies regarding the classification of online ojek service review using the Naïve Bayes algorithm, C.45 and Random Forest produced an unsatisfactory accuracy of 69.18% at the highest value. This study aims to determine the extent of the analysis of Gojek and Grab application user reviews based on user comments by classifying negative and positive reviews with a higher level of accuracy than previous studies so that applications with the best reviews can be known for public consideration in using the application's services. The method used for data review classification is using the Support Vector Machine (SVM) based on Particle Swarm Optimization (PSO). The test results on the Grab application review get the highest accuracy results in the amount of 73.09% with AUC value = 0.804, while for the test results on the application review Gojek get an accuracy value of 65.59% and AUC value = 0.680</p> Hermanto Hermanto Antonius Yadi Kuntoro Taufik Asra Nurajijah Nurajijah Lasman Effendi Ridatu Ocanitra ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 117 122 10.33480/pilar.v16i1.1304 SENTIMENT ANALYSIS ON CLOSURE OF ILLEGAL MOVIE STREAMING SITES USING NAÏVE BAYES ALGORITHM https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1306 <p>The closure of illegal movie streaming sites IndoXXI has been a trending topic on Twitter at the end of 2019. The reaction of netizens on Twitter shows positive and negative sentiments. Until now, there have been many studies in the field of Sentiment Analysis using data in the form of Tweets from Twitter users. In sentiment analysis research, there are so many method used, and Naïve Bayes is one of it, because it is very simple and efficient. The method has advantages and disadvantages. Naïve Bayes is so sensitive in feature selection. Too many features not only increase calculation time but also reduce classification accuracy. In order to solve the disadvantages and increase the performance of the Naïve Bayes classifier, this method often being combined with many kind of feature selection methods. This research aims to classify tweets into positive and negative using the Naïve Bayes classifier combined with the Genetic Algorithm. The accuracy of Naïve Bayes before using the combination of feature selection methods reaches 79.55%. While after using feature selection methods, which is the Genetic Algorithm, accuracy increased up to 88.64%. The accuracy improved by up to 9.09%.</p> Dinda Ayu Muthia ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 123 128 10.33480/pilar.v16i1.1306 NEURAL NETWORK OPTIMIZATION WITH PARTICLE SWARM OPTIMIZATION AND BAGGING METHODS ON CLASSIFICATION OF SINGLE PAP SMEAR IMAGE CELLS https://ejournal.nusamandiri.ac.id/index.php/pilar/article/view/1308 <p>In this study, an automatic diagnosis analysis of the results of pap smear image extraction using neural network algorithms, the analysis included a review of the results of Herlev pap smear extraction level 7 grade, 2 normal and abnormal classes, 3 classes of normal level dysplasia and 4 classes of abnormal dysplasia levels. The problem is that neural networks are very difficult to designate optimal features in diagnosing and difficult to handle class imbalances. This study proposes a combination of particle swarm optimization (PSO) to optimize the features and bagging methods to deal with class imbalances, with the aim that the results of diagnosis using a neural network can increase its accuracy. The results show that using PSO and bagging methods can improve the accuracy of the algorithm of network balance. At level 7 the buffer class increased by 1.64%, 2 classes increased by 0.44%, 3 classes increased by 2.04%, and at level 4 the class increased by 5.47%In this study, an automatic diagnosis analysis of the results of pap smear image extraction using neural network algorithms, the analysis included a review of the results of Herlev pap smear extraction level 7 grade, 2 normal and abnormal classes, 3 classes of normal level dysplasia and 4 classes of abnormal dysplasia levels. The problem is that neural networks are very difficult to designate optimal features in diagnosing and difficult to handle class imbalances. This study proposes a combination of particle swarm optimization (PSO) to optimize the features and bagging methods to deal with class imbalances, with the aim that the results of diagnosis using a neural network can increase its accuracy. The results show that using PSO and bagging methods can improve the accuracy of the algorithm of network balance. At level 7 the buffer class increased by 1.64%, 2 classes increased by 0.44%, 3 classes increased by 2.04%, and at level 4 the class increased by 5.47%</p> Robi Aziz Zuama Irwan Agus Sobari ##submission.copyrightStatement## http://creativecommons.org/licenses/by-nc/4.0 2020-03-31 2020-03-31 16 1 129 134 10.33480/pilar.v16i1.1308