Jurnal KomtekInfo

Evaluation of E-Government Governance with the Implementation of the COBIT 2019 Method at the West Pasaman Civil Registration Office

Edo Septiawan, Jhon Veri, S Sumijan — Tue, 30 Dec 2025 00:00:00 +0700

The rapid development of information technology has encouraged government agencies to adopt Electronic Government (E-Government) to improve the efficiency, transparency, and accountability of public services. However, implementation challenges remain, especially in regional agencies such as the Population and Civil Registration Office (Dukcapil) of West Pasaman Regency, where inconsistencies and lack of system integration are still widespread. This study aims to analyze and evaluate E-Government governance. The method used in this study is the COBIT 2019 framework to assess the level of capability and propose strategic improvements in IT service management. This study adopted a qualitative case study approach, guided by the COBIT 2019 Design and Implementation framework. Data were collected through direct observation and in-depth interviews with Dukcapil personnel, supported by 145 structured questions and documentary analysis. This process focuses on five COBIT processes: EDM04 (Ensure Resource Optimization), APO07 (Manage Human Resources), BAI09 (Manage Assets), DSS01 (Manage Operations), and MEA01 (Monitor, Evaluate, and Assess Performance and Conformance). The results show that the capability level of the DSS01 process is at level 2 (Managed Process), while the capability target is level 3 (Established Process). This gap reflects the need for improvement in operational process management, including strengthening documentation and standardizing practices. Based on these results, strategic recommendations are developed that are contextually oriented to local bureaucratic conditions. This research can be a reference for the implementation of COBIT 2019 to effectively identify governance weaknesses and provide actionable strategies in increasing E-Government maturity at the regional level, supporting better service delivery and organizational performance.

Convolutional Neural Network Architecture Densenet121 to Identify Tuberculosis

Fajri Nugraha, Sumijan S, Rini Sovia — Tue, 30 Dec 2025 00:00:00 +0700

Smoking habits and the normalization of smoking activities are often a problem in many developing countries in the world. Cigarette smoke can cause many health problems that increase the risk of developing diseases and worsen the condition of people with the disease, one of which is Tuberculosis (TB). In Indonesia, based on the WHO Global TB Report 2024, Indonesia ranks second in the world in TB cases, it is estimated that there are more than 1,000,000 new cases every year, this disease is a very serious health problem and has obstacles in the identification process. This research aims to develop a TB disease identification system using Deep Learning. The methods used in this study are Convolutional Neural Network (CNN) and Densenet121 architecture. Convolutional Neural Network (CNN) was chosen for its ability to perform X-ray image analysis for visual validation, while Densenet121 was chosen because of its flexible architecture that can be applied to a wide range of computer vision applications, including image classification, object identification, and semantic segmentation. The research stage includes data collection, then preprocessing the image, namely resize, normalization, and conversion to arrays, then building a Convolutional Neural Network model with the selected architecture, then model training, model performance evaluation using accuracy and AUC metrics and ending with testing and validation by experts. The dataset used in this study is X-Ray data of tuberculosis patients taken from Kaggle to build a Deep Learning model that is able to identify TB through 100 chest X-ray image datasets. The results of the study show that the CNN model is able to identify tuberculosis with an accuracy rate of up to 90%, so it can help speed up early diagnosis or screening so that patients can continue to receive treatment and treatment. Therefore, the application of deep learning with the Convolutional Neural Network (CNN) method and DenseNet121 architecture based on X-Ray image data is an effective approach in the early detection of tuberculosis and seeks to make an important contribution to the control of lung diseases related to exposure to cigarette smoke in Indonesia.

Optimization of LPG Gas Distribution Routes with a Combination of the Saving Matrix Method and Nearest Neighbor

Andi Amin Amirul Mukminin, Billy Hendrik, Rini Sovia — Tue, 30 Dec 2025 00:00:00 +0700

Distribution is an important process in economic activities, which involves the delivery of goods or products from producers to end consumers. Efficiency in the distribution system highly depends on the selection of optimal routes, which can affect costs, time, and the quality of service provided. PT Amartha Anugrah Mandiri, which operates in the distribution of 3 kg LPG, faces significant challenges in terms of inefficient distribution route selection, limited fleet capacity, and unstructured variations in LPG demand. The distribution routes currently used do not consider the aspects of distance, time, and cost efficiency, resulting in the wastage of resources such as fuel and time. This research aims to optimize LPG distribution routes. The methods used in this study are the Saving Matrix and Nearest Neighbor. The Saving Matrix method is used to reduce distribution distance and costs by combining existing delivery routes, while the Nearest Neighbor is applied to determine the order of visits to the nearest bases gradually. Both methods are designed to produce distribution routes that are efficient in terms of time, distance, and cost, as well as to maximize the use of the existing fleet. The data in this study were obtained thru direct observation at PT. Amartha Anugrah Mandiri. The data collected included base locations, LPG demand, vehicle capacity, and operational costs. There are 22 bases served with a total delivery reaching 1120 LPG 3 kg cylinders spread across various sub-districts of Batam City. Deliveries are carried out using trucks with a maximum capacity of 560 cylinders, so in one day, distribution requires more than one trip. Using this data, the distance matrix and savings matrix were calculated to design a more efficient distribution system. The research results show that the application of these two methods successfully reduced the total distance traveled, delivery time, and operational costs significantly, as well as improved the efficiency of LPG distribution. This research is expected to contribute to the company so that the 3 kg LPG delivery process can run optimally.

Sentiment Analysis in Platform X with the Support Vector Machine Method for Generation Z

Apriandini Sri Dewi, Sarjon Defit, Gunadi Widi Nurcahyo — Tue, 30 Dec 2025 00:00:00 +0700

Advances in information technology and the increasing use of social media have significantly influenced the behavior of Generation Z. The generation born between 1997 and 2012 is known to be very familiar with the digital world, but also faces challenges such as lack of in-person social interaction and the risk of mental health disorders. This study aims to identify and classify public sentiment towards Generation Z on social media, especially on platform X (formerly Twitter). The method used is the Support Vector Machine (SVM). This research was carried out through several stages, namely the collection of 1607 data in the form of text using crawling techniques, pre-processing of text (tokenization, case folding, removal of stopwords, stemming, and normalization), and feature extraction using the Term Frequency-Inverse Document Frequency (TF-IDF) method. The processed data is then classified into three sentiment categories: positive, negative, and neutral using SVM. Evaluation was carried out by measuring accuracy, recall value, and F1-score value through a confusion matrix. The results showed that the measurement of an accuracy value of 85%, a precision value of 85%, a value of recall of 95% and an F1-score value of 90% that SVM was able to classify sentiment with high accuracy and stability. In addition, SVM has been shown to be more effective than other methods studied in previous studies. The data analyzed shows that most sentiment towards Generation Z is negative, reflecting public concern about the behavior and mindset of this generation. This research is expected to be a reference for academics, practitioners, and policymakers in understanding public opinion and designing targeted policies for the younger generation.

Keywords: Sentiment Analysis, Generation Z, Support Vector Machine, Social Media, Machine Learning.

Convolutional Neural Network Method in Detecting Digital Image Based Physical Violence

Elpina Sari Dewi Hasibuan Elpina, Y Yuhandri, S Sumijan — Tue, 30 Dec 2025 00:00:00 +0700

Physical violence in the educational environment has a serious impact on mental health, safety, and student achievement, in addition to causing physical injury, violence can cause psychological trauma that interferes with the learning process, due to the limited supervision system, lack of officers, and the absence of automatic detection technology. This research aims to design and develop an automatic detection system of physical violence using digital image processing technology. This study uses the Convolutional Neural Network (CNN) method with the stages of digital image collection and labeling, preprocessing, model training, and evaluation using accuracy, precision, recall, and F1-score metrics. The CNN architecture was chosen because it is efficient and accurate, and it supports data augmentation to improve generalization. The dataset was taken from kaggle and primary data at the al-falah huraba Islamic boarding school which consisted of 2000 images which included: 800 images of violence on CCTV of the dormitory room, 500 images of violence simulation of training videos and 500 non-violent images. The results showed that the developed CNN model was able to detect physical violence with an accuracy of above 88%, making it feasible to apply in surveillance camera-based school surveillance systems (CCTV). The system is able to classify images in real-time into two categories: safe and hard. This research contributes to the use of artificial intelligence to support efficient and affordable technology-based education security.

Combination of Support Vector Machine and Artificial Neural Network Methods in Negative Content Filtering System

M Wira Sanjaya Wira; Y Yuhandri; Billy Hendrik — Tue, 30 Dec 2025 00:00:00 +0700

Local Wi-Fi network access has become a common necessity in everyday digital activities, but it is vulnerable to misuse to access negative content. This content includes pornographic material, hate speech, and violent content that can adversely affect users, especially in educational settings. For this reason, a system that is able to filter malicious content automatically and efficiently is needed. This research aims to design an artificial intelligence-based negative content filtering system that can be run on local network devices. The methods used include image classification using Convolutional Neural Network (CNN) and Artificial Neural Network (ANN), as well as text classification with DistilBERT and Support Vector Machine (SVM). To maintain user privacy, the model is trained using a federated learning approach that allows for decentralized learning. Knowledge distillation is also applied to produce lightweight models that can be run on edge devices such as routers. The datasets used include NSFW Image Dataset, OpenPornSet, as well as a collection of toxic comments from Reddit and Twitter. The evaluation was carried out in a simulation of a local network with 50 active devices. The test results showed an ANN accuracy rate of 93.4% in recognizing visual content, and SVM accuracy of 91.7% in detecting text-based hate speech. This research can be a reference in the application of AI-based content filtering systems for safe and responsible digital access protection

Decision Support System in Determining TPQ/TQA Teacher Certification Categories Using the SAW Method

Afdal Zikri, Gunadi Widi Nurcahyo, S Sumijan — Tue, 30 Dec 2025 00:00:00 +0700

TPQ/TQA teacher certification is an effort to improve the quality of educators in deepening their knowledge of the Qur'an. The certification assessment process often faces challenges related to subjectivity and inconsistencies in criteria, thus requiring a decision support system capable of producing more objective and measurable assessment results. Based on the problems described above, this study aims to analyze the TPQ/TQA teacher certification assessment in Padang City. The SAW method is very suitable for this study because of its ability to perform calculations based on predetermined criteria. The research data consists of 60 assessment documents. The analysis process includes determining criteria, normalizing weights, calculations, and rankings. Based on the 60 datasets, 9 individuals obtained a certification score of A, 11 obtained a B, and 40 obtained a C. The results of this study indicate that the decision support system is capable of providing highly accurate, transparent, and efficient results in determining TPQ/TQA teacher certification scores. These findings are expected to be useful for TPQ/TQA management institutions in determining certification scores.

Comparison of Random Forest and Support Vector Machine Learning Algorithms in Sentiment Analysis of Gojek User Reviews

Tesa Vausia Sandiva, Arip Kristiyanto — Tue, 30 Dec 2025 00:00:00 +0700

The development of digital technology has brought significant changes across various sectors of life, including transportation. One of the most popular modes of transportation among the public today is online motorcycle taxis, such as Gojek. Gojek continues to innovate to meet customer needs more effectively and to expand its range of services. This study aims to identify the number of positive, neutral, and negative sentiments in a user review dataset, as well as to evaluate the performance of the algorithms used—namely, SVM and Random Forest. The analysis was conducted on 10,000 customer reviews from the Play Store application, resulting in 2,057 positive sentiments, 1,135 neutral sentiments, and 6,295 negative sentiments. The classification model compared the SVM algorithm with the Random Forest algorithm, and the results show that Random Forest achieved better performance, with 91% accuracy compared to SVM’s 89%. These findings demonstrate that Random Forest performs better in handling word distribution within review texts than the SVM method.

Combination of Active Contour and CNN-based Segmentation Methods to Improve Accuracy in Detecting Rice Diseases

Wahyu Saptha Negoro, Ratih Adinda Destari, Asbon Hendra Azhar, Achmad Syahrian — Tue, 30 Dec 2025 00:00:00 +0700

Rice diseases are one of the main factors causing decreased productivity and threatening national food security. The main problem in controlling rice diseases is the delay and inaccuracy of symptom identification in the field. This study aims to develop an artificial intelligence-based rice disease detection system through a combination of Active Contour and Convolutional Neural Network (CNN) methods. The research object is rice leaf images taken from rice fields in Pulau Sejuk Village, Batubara Medan, with a dataset of 600 images consisting of healthy leaves and 3 types of rice diseases. The Active Contour method is used in the segmentation stage to extract leaf areas precisely, while CNN is applied for the disease classification process. The results show that this combination of methods can significantly improve the accuracy of rice disease detection. The developed system is expected to assist farmers and stakeholders in the early detection of rice diseases, thereby supporting food innovation and increasing sustainable agricultural productivity.