Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi)
Not a member yet
1071 research outputs found
Sort by
Exploring Research Trends and Impact: A Bibliometric Analysis of RESTI Journal from 2018 to 2022
Get PDFThis study provides a comprehensive analysis of the RESTI Journal, a prominent publication in the field of systems engineering and information technology. The analysis aims to evaluate the journal's publication output, citation impact, and overall contribution to the field. The study utilizes data from the Dimensions database, focusing on articles published between 2018 and 2022, resulting in a dataset of 594 articles. To analyze the collected data, the study employs bibliometric and network visualization tools such as Bibliometrix and VOSviewer. The analysis reveals a notable increase in the number of publications over time, indicating a growing interest and research activity in the field. Furthermore, the distribution of author productivity deviates from Lotka's law, highlighting variations in author patterns and productivity levels. An examination of institutional affiliations reveals Telkom University as the dominant institution, making a substantial contribution to the journal. Visualizations based on author-provided titles, abstracts, and keywords highlight research trends in image recognition and classification, with a particular emphasis on utilizing Convolutional Neural Networks (CNN) and Support Vector Machines (SVM). Overall, this study provides valuable insights into the performance and trends of the RESTI Journal. The findings contribute to a deeper understanding of the journal's impact and its role in advancing knowledge in systems engineering and information technology. These insights can inform researchers, practitioners, and stakeholders in the field, guiding future research directions and enhancing the scholarly impact of the RESTI Journal
Comparative Analysis of Support Vector Machine and Perceptron In The Classification of Subsidized Fuel Receipts
Get PDFCurrently, fuel oil is one of the important factors for the community and even a country on this earth to utilize this natural gas fuel for daily use as the main use and also by increasing the community's need for fuel oil. But there are several factors that cause this fuel problem, there is a factor of time and usage time, which is certain that one day it will expire and its capacity in a country, even if the country runs out of fuel, will make requests to other countries and also obstacles to supplying this fuel oil to the public. which is the main fuel from the Pertamina government agency which has begun to limit purchases for this fuel oil to certain circles by marking the types of subsidies or not subsidies that must be controlled by the government in limiting purchases for the public. In dealing with solving problems from the perspective of ownership or even utilization, there are limits to owning fuel, and not everyone has to have a lot or even too much. In solving the problem of dividing fuel revenue, which is good for filling revenue, it can be solved by using machine learning, namely data mining itself can help in completing subsidized fuel receipts without being excessive for the community so that they can be controlled and managed for their purchases. In building a fuel oil reception design, it can be grouped into a classification model that uses SVM and perceptron which uses the activation function of the sigmoid to get the final result of accuracy where getting the average value of 5-fold, 10-fold, 20-fold is accuracy. is 90.0%, the F1 value is 85.6%, the precision value is 87.6%, and the recall value is 90.0%.Currently, fuel oil is one of the important factors for the community and even a country on this earth to utilize this natural gas fuel for daily use as the main use and also by increasing the community's need for fuel oil. But there are several factors that cause this fuel problem, there is a factor of time and usage time, which is certain that one day it will expire and its capacity in a country, even if the country runs out of fuel, will make requests to other countries and also obstacles to supplying this fuel oil to the public. which is the main fuel from the Pertamina government agency which has begun to limit purchases for this fuel oil to certain circles by marking the types of subsidies or not subsidies that must be controlled by the government in limiting purchases for the public. In dealing with solving problems from the perspective of ownership or even utilization, there are limits to owning fuel, and not everyone has to have a lot or even too much. In solving the problem of dividing fuel revenue, which is good for filling revenue, it can be solved by using machine learning, namely data mining itself can help in completing subsidized fuel receipts without being excessive for the community so that they can be controlled and managed for their purchases. In building a fuel oil reception design, it can be grouped into a classification model that uses SVM and perceptron which uses the activation function of the sigmoid to get the final result of accuracy where getting the average value of 5-fold, 10-fold, 20-fold is accuracy. is 90.0%, the F1 value is 85.6%, the precision value is 87.6%, and the recall value is 90.0%
Sentiment Analysis of Cryptocurrency Trading Platform Service Quality on Playstore Data: A Case of Indodax
Get PDFIndodax is one of the cryptocurrency trading platforms in Indonesia that has the highest sentiment for the quality they provide, good quality on a platform is an important factor in obtaining user satisfaction and will have an impact on the long-term success of a company. The importance of user satisfaction on cryptocurrency online trading platforms is a significant factor in increasing user loyalty in today's competition. This research was conducted to analyze the quality of existing cryptocurrency trading platform services so that they can be input for cryptocurrency trading service providers to improve the quality of their services, this information can also be considered by prospective platform users in choosing a trading platform that has the best quality of service to minimize losses that may be caused by the platform. In this study, sentiment analysis was used for indodax play store platform users and then processed using the lexicon classification method to produce sentiment analysis for each significant factor of service quality. From the results of the classification carried out in this study, the results of the analysis show that most users are satisfied and give positive sentiments related to security, namely 87.63%, positive sentiments related to the interface design 88.46%, positive sentiments related to service & convenience by 83%, but some users also gave a slightly positive sentiment related to administrative costs, namely 39%, and their negative sentiment was mostly related to the error & failure system, which received more than 80% sentiment. While the recall value is 38.07%, the precision is 56.69% and the f1-score is 45.55%. The results of this study can be concluded that there are still many important points that must be improved in quality by the indodax platform service providers so that they can be more attractive and used by everyone.
Indodax is one of the cryptocurrency trading platforms in Indonesia that has the highest sentiment for the quality they provide, good quality on a platform is an important factor in obtaining user satisfaction and will have an impact on the long-term success of a company. The importance of user satisfaction on cryptocurrency online trading platforms is a significant factor in increasing user loyalty in today's competition. This research was conducted to analyze the quality of existing cryptocurrency trading platform services so that they can be input for cryptocurrency trading service providers to improve the quality of their services, this information can also be considered by prospective platform users in choosing a trading platform that has the best quality of service to minimize losses that may be caused by the platform. In this study, sentiment analysis was used for indodax play store platform users and then processed using the lexicon classification method to produce sentiment analysis for each significant factor of service quality. From the results of the classification carried out in this study, the results of the analysis show that most users are satisfied and give positive sentiments related to security, namely 87.63%, positive sentiments related to the interface design 88.46%, positive sentiments related to service & convenience by 83%, but some users also gave a slightly positive sentiment related to administrative costs, namely 39%, and their negative sentiment was mostly related to the error & failure system, which received more than 80% sentiment. While the recall value is 38.07%, the precision is 56.69% and the f1-score is 45.55%. The results of this study can be concluded that there are still many important points that must be improved in quality by the indodax platform service providers so that they can be more attractive and used by everyone
Pedestrian Detection System using YOLOv5 for Advanced Driver Assistance System (ADAS)
Get PDFThe technology in transportation is continuously developing due to reaching the self-driving vehicle. The need of detecting the situation around vehicles is a must to prevent accidents. It is not only limited to the conventional vehicle in which accident commonly happens, but also to the autonomous vehicle. In this paper, we proposed a detection system for recognizing pedestrians using a camera and minicomputer. The approach of pedestrian detection is applied using object detection method (YOLOv5) which is based on the Convolutional Neural Network. The model that we proposed in this paper is trained using numerous epochs to find the optimum training configuration for detecting pedestrians. The lowest value of object and bounding box loss is found when it is trained using 2000 epochs, but it needs at least 3 hours to build the model. Meanwhile, the optimum model’s configuration is trained using 1000 epochs which has the biggest object (1.49 points) and moderate bounding box (1.5 points) loss reduction compared to the other number of epochs. This proposed system is implemented using Raspberry Pi4 and a monocular camera and it is only able to detect objects for 0.9 frames for each second. As further development, an advanced computing device is needed due to reach real-time pedestrian detection.
The technology in transportation is continuously developing due to reaching the self-driving vehicle. The need of detecting the situation around vehicles is a must to prevent accidents. It is not only limited to the conventional vehicle in which accident commonly happens, but also to the autonomous vehicle. In this paper, we proposed a detection system for recognizing pedestrians using a camera and minicomputer. The approach of pedestrian detection is applied using object detection method (YOLOv5) which is based on the Convolutional Neural Network. The model that we proposed in this paper is trained using numerous epochs to find the optimum training configuration for detecting pedestrians. The lowest value of object and bounding box loss is found when it is trained using 2000 epochs, but it needs at least 3 hours to build the model. Meanwhile, the optimum model’s configuration is trained using 1000 epochs which has the biggest object (1.49 points) and moderate bounding box (1.5 points) loss reduction compared to the other number of epochs. This proposed system is implemented using Raspberry Pi4 and a monocular camera and it is only able to detect objects for 0.9 frames for each second. As further development, an advanced computing device is needed due to reach real-time pedestrian detection
Detecting Alter Ego Accounts using Social Media Mining
Get PDFAlter ego is a condition of someone who creates a new character with a conscious state. Original character role play is a game to create new imaginary characters that is used as research material for identification alter ego accounts. The negative effects of playing alter ego are stress, depression, and multiple personalities. Current research only focuses on the phenomenon and impacts of a role-playing game. We propose a new method to detect accounts of alter ego players in social media, especially Twitter. We develop an application to analyze the characteristics of alter ego accounts. Psychologists can use this application to discover the characteristics of alter ego accounts that are useful for analyzing personality so that the results can be used to appropriately handle alter ego players. Most user profiles, tweets, and platforms are used to detect account Twitter. This research proposes a new method using bio features as input data. We crawled and collected 565 bios from Twitter for one month. We observe the data to search for unique words and collect them into a classification dictionary. In this research, we use the cosine similarity method because this method is popular for detecting text and has a good performance in many cases. This research could identify alter ego accounts and other types of Twitter accounts. From the detection results of alter ego accounts, it is possible to analyze the characteristics of Twitter accounts. We use a sampling technique that takes 30% of the data as testing data. According to the results of the experiment cosine similarity obtained an accuracy of 0.95.
Alter ego is a condition of someone who creates a new character with a conscious state. Original character role play is a game to create new imaginary characters that is used as research material for identification alter ego accounts. The negative effects of playing alter ego are stress, depression, and multiple personalities. Current research only focuses on the phenomenon and impacts of a role-playing game. We propose a new method to detect accounts of alter ego players in social media, especially Twitter. We develop an application to analyze the characteristics of alter ego accounts. Psychologists can use this application to discover the characteristics of alter ego accounts that are useful for analyzing personality so that the results can be used to appropriately handle alter ego players. Most user profiles, tweets, and platforms are used to detect account Twitter. This research proposes a new method using bio features as input data. We crawled and collected 565 bios from Twitter for one month. We observe the data to search for unique words and collect them into a classification dictionary. In this research, we use the cosine similarity method because this method is popular for detecting text and has a good performance in many cases. This research could identify alter ego accounts and other types of Twitter accounts. From the detection results of alter ego accounts, it is possible to analyze the characteristics of Twitter accounts. We use a sampling technique that takes 30% of the data as testing data. According to the results of the experiment cosine similarity obtained an accuracy of 0.95
Triangular Fuzzy Numbers-Based MADM for Selecting Pregnant Mothers at Risk of Stunting
Get PDFStunting is caused by a lack of proper nutrition before and after birth. This research paper identifies and measures the risk of stunting during pregnancy and make recommendations for ranking pregnant women at risk. These aims to provide appropriate treatment and action to reduce mothers giving birth to children at risk of stunting. To make the optimal choice, the selection procedure for pregnant women at risk of giving birth to stunted children considers a variety of factors, including maternal age, maternal nutrition, arms circumference, hemoglobin, parity, birth interval, height, baby weight, and body mass index (BMI). Decision-maker’s expectation to reduce uncertainty and imprecision are represented linguistically by triangular fuzzy numbers. The triangular fuzzy numbers arithmetic approach is used to determine the selection process output. The ranking is determined from the alternative with the most parameter values to the alternative with the fewest parameters. Based on the results of the calculation, it was determined that PM (Pregnant Mother) had the highest score and was ranked first. That pregnant mother was declared as pregnant mother who had the lowest risk of giving birth to stunted baby
Stunting is caused by a lack of proper nutrition before and after birth. This research paper identifies and measures the risk of stunting during pregnancy and make recommendations for ranking pregnant women at risk. These aims to provide appropriate treatment and action to reduce mothers giving birth to children at risk of stunting. To make the optimal choice, the selection procedure for pregnant women at risk of giving birth to stunted children considers a variety of factors, including maternal age, maternal nutrition, arms circumference, hemoglobin, parity, birth interval, height, baby weight, and body mass index (BMI). Decision-maker’s expectation to reduce uncertainty and imprecision are represented linguistically by triangular fuzzy numbers. The triangular fuzzy numbers arithmetic approach is used to determine the selection process output. The ranking is determined from the alternative with the most parameter values to the alternative with the fewest parameters. Based on the results of the calculation, it was determined that PM (Pregnant Mother) had the highest score and was ranked first. That pregnant mother was declared as pregnant mother who had the lowest risk of giving birth to stunted bab
Sentiment Analysis of Electricity Company Service Quality Using Naïve Bayes
Get PDFIn facing the era of technological disruption, a large company providing electricity in Indonesia, namely PT PLN is transforming to digitize all business processes and improve the quality of customer service. PLN Mobile application was developed in December 2020, and 18 million users have downloaded it. PLN Mobile application provides various electrical services for users. There are a lot of online opinions today. Organizations need to know the public perception of their product or service, sales projections, and customer happiness. Our research will identify public opinion (positive and negative) about PLN Mobile Application using sentiment analysis by taking review data from Google Play Store. Sentiment analysis is classified using Naïve Bayes and analyzed based on the dimensions of the quality of electricity services: empathy, responsiveness, and reliability. The results of this study indicate that Naïve Bayes is quite well used for binomial labels (positive and negative) with an accuracy of 73%. Still, for service quality dimensions, the accuracy is 45%. Indonesian language datasets are quite difficult to process due to non-standard language, foreign words, mixed language variations, and abbreviations. Determination of ground truth or manual labeling requires consistency and skilled personnel to determine the context of the text data to obtain a model with optimal performance. This study informs the classification of each dimension of the quality of electricity services in Indonesia based on positive and negative sentiment data for PLN Mobile Application users. Reliability received the most negative sentiments. This can be used for PT PLN to improve the quality-of-service reliability to customers.
In facing the era of technological disruption, a large company providing electricity in Indonesia, namely PT PLN is transforming to digitize all business processes and improve the quality of customer service. PLN Mobile application was developed in December 2020, and 18 million users have downloaded it. PLN Mobile application provides various electrical services for users. There are a lot of online opinions today. Organizations need to know the public perception of their product or service, sales projections, and customer happiness. Our research will identify public opinion (positive and negative) about PLN Mobile Application using sentiment analysis by taking review data from Google Play Store. Sentiment analysis is classified using Naïve Bayes and analyzed based on the dimensions of the quality of electricity services: empathy, responsiveness, and reliability. The results of this study indicate that Naïve Bayes is quite well used for binomial labels (positive and negative) with an accuracy of 73%. Still, for service quality dimensions, the accuracy is 45%. Indonesian language datasets are quite difficult to process due to non-standard language, foreign words, mixed language variations, and abbreviations. Determination of ground truth or manual labeling requires consistency and skilled personnel to determine the context of the text data to obtain a model with optimal performance. This study informs the classification of each dimension of the quality of electricity services in Indonesia based on positive and negative sentiment data for PLN Mobile Application users. Reliability received the most negative sentiments. This can be used for PT PLN to improve the quality-of-service reliability to customers
Robust Digital Watermarking pada Arsip Vital Mnggunakan Metode Hybrid SVD Dengan DWT
No full textThe development of Internet technology affects the dissemination of data, especially in vital government archives. This research uses a hybrid singular value decomposition (SVD) and discrete wavelet transform (DWT) method, which aims to protect the copyright of vital archives. The stages of the insertion and extraction process are carried out to test the effect of the alpha value on the quality (imperceptibility) and robustness of the inserted image by measuring the Peak Signal-to-Noise Ratio (PSNR), verifying similarity by measuring the Normalized Cross-Correlation (NC) and Structural Similarity Index (SSIM). The results of research with ten vital archives and a watermark protection logo in JPEG format with a size of 512x512 pixels obtained a maximum PSNR with a value of α = 0.01 of 41.0567 dB, NC of 0.98904, and SSIM of 0.98023 in the Cibereum Land Certificate. So, it can be proven that this method produces vital archive watermarks that can be extracted and are robust to JPEG compression attacks of 75%, median filtering 3x3, Gaussian noise 0.01, speckle noise 0.01, and salt and pepper noise 0.01 but not resistant to rotation 80 and cropping attacks 2%.Perkembangan teknologi industri 4.0 membuat keamanan hak cipta dan kekayaan intelektual menjadi suatu masalah yang sangat penting untuk perkembangan teknologi multimedia, seperti text, image, graphic, audio, dan video dimana hak kepemilikan atau copyright label perlu dilakukan perlindungan. Pada penelitian ini akan digunakan metode digital watermarking untuk melindungi copyright arsip vital pada pemerintahan menggunakan hybrid Singular Value Decomposition (SVD) dengan Discrete Wavelet Transform (DWT). Penelitian ini mengusulkan kedua metode tersebut untuk melakukan pengujian kualitas citra hasil penyisipan dan ketahanan terhadap serangan dari penyalahgunaan copyright dengan melihat nilai alpha yang akan mempengaruhi imperceptibility dan robustness citra arsip vital. Hasil pengujian metode penelitian tanpa serangan dengan nilai α=0.01 mendapatkan nilai Peak Signal to Noise Ratio (PSNR) terbaik sebesar 51.7825 dB, Normalized Cross-Correlation (NC) sebesar 0.9995, dan Structural Similarity Index (SSIM) sebesar 0.99843 pada Sertifikat Tanah Cibereum. Hasil dari pengujian pada serangan kompresi JPEG dengan quality 75% didapatkan nilai PSNR terbaik pada Sertifikat Tanah Cibereum dengan nilai PSNR sebesar 41.0567 dB, NC sebesar 0.98904, dan SSIM sebesar 0.98023. Pada hasil pengujian filtering median 3x3 citra dapat diekstraksi dengan baik menggunakan nilai α=0.01 didapatkan nilai PSNR terbaik sebesar 35.0168 dB, NC sebesar 0.942, dan SSIM sebesar 0.96965 yaitu pada Sertifikat Tanah Gunung Sindur. Jadi dapat dibuktikan metode hybrid SVD dengan DWT sangat baik digunakan untuk proteksi pada citra digital arsip vital untuk menjaga autentifikasi hak kepemilikan citra, dan lebih tahan (robust) terhadap berbagai serangan
Service Automation Implementation for Delivering CaaS at the Ministry of Finance of Indonesia
Get PDFEvaluation is an essential aspect of service improvement. Within the Ministry of Finance, there is an organization responsible for providing IT services to various units and employees. One of the services offered by this organization is cloud computing, which supports the development of information systems. However, there are several challenges related to service fulfillment. For instance, the time required to fulfill a service is relatively long, taking two days, in contrast to public cloud providers, which can deliver their services in minutes. Additionally, there is a potential for human errors in the manual process carried out by the Request Fulfillment Team (RFT) during service delivery. This study aims to explore the design and implementation of automated container service fulfillment, transforming them into self-service products. The author employs the Finite-State Automata (FSA) model to test the input and output of the automation system using seven states and inputs. The results indicate that the container service cycle, when compiled and tested with FSA using predetermined inputs, can generate containers according to user-selected specifications. As a result, the implementation of an automated and self-service model is proposed to reduce delivery time and mitigate potential errors in the container-as-a-service (CaaS) offering
Optimasi Hyperparameter dari CNN Classifier untuk Klasifikasi Genre Musik
No full textPlaying music through a digital platform that has a large database of songs requires automated classification of music genres, highlighting the need to develop a model for music genre classification that is more efficient and accurate. This study evaluated the hyperparameters in the music genre classification process using CNN in the GTZAN dataset with 30-second duration data optimized using MFCC feature extraction. The model that is formed with a time of 3 (three) seconds classifies music genres in the first 3 seconds of music. This model has a high potential for error because the first 3 seconds of initial music are varied and cannot be used as a benchmark in determining music genres. This study performed hyperparameters on batch size, epoch, and split data set variables with various scenarios. The highest precision result was obtained at 72% with a data split of 85%:15%, 32 batch sizes, and 500 epochs.Memutar musik melalui platform digital yang memiliki database lagu yang besar membutuhkan klasifikasi genre musik yang berjalan secara otomatis, sehingga penelitian mengenai klasifikasi genre musik menjadi penting dan dicari metode yang lebih efisien dan akurat. Studi ini mengevaluasi hyperparameter dalam proses klasifikasi genre musik menggunakan Convolution Neural Network pada dataset GTZAN dengan data durasi 30 detik yang dioptimalkan menggunakan ekstraksi fitur MFCC. Model yang dibentuk dengan waktu 3 (tiga) detik ini akan mengklasifikasikan genre musik pada 3 detik pertama musik yang memiliki potensi error yang tinggi karena musik awal terutama 3 detik sangat variatif dan tidak dapat digunakan sebagai patokan dalam menentukan genre musik. Evaluasi ini melakukan hyperparameter pada variabel batch size, epoch dan split dataset dengan berbagai skenario. Hasil akurasi tertinggi diperoleh sebesar 72% dengan data split 85%-15%, ukuran batch 32 dan 500 epoc