Jurnal Sistemasi (OJS FTIK - UNISI, Fakultas Teknik dan Ilmu Komputer Universitas Islam Indragiri)
Not a member yet
946 research outputs found
Sort by
School Profile Website using the K-Means Algorithm at the Minahasa Regency Education Office
Full text linkBackground: The Education Office often manages school profiling with minimal and fragmented data, such as school name, district, level, total students, and total teachers. This situation makes planning reactive and error-prone, highlighting the need for a lightweight yet reliable workflow to transform aggregated data into actionable evidence. Method: This study developed a web-based profiling tool that integrates a user interface designed using a User-Centered Design (UCD) approach with a transparent K-Means clustering algorithm. Development was carried out through iterative prototyping, with features standardized using z-scores and cluster validity assessed using the silhouette method along with other internal validity indices. Results: Data entry features with header previews and numeric checks effectively reduced rework. The silhouette value peaked at , producing two interpretable segments (moderate vs. high staffing load), with an optional for exploratory purposes. Usability evaluation using the System Usability Scale (SUS) yielded a score of approximately 82, indicating good user acceptance, with system response times measured in seconds.
Conclusion: The system provides a streamlined and sufficient workflow for routine planning and establishes a foundation for future longitudinal developments, such as tracking the number of study groups and accreditation per period
Implementation of a Waste Bank Website using the Agile Scrum Approach and User Acceptance Testing (UAT)
Full text linkWaste management in Indonesia continues to face significant challenges, largely due to manual record-keeping practices that often lead to data duplication, reporting delays, and a lack of transparency. This study develops EcoHub, an integrated digital dashboard designed to monitor waste deposit flows, point conversion, and product transactions to support a circular economy ecosystem. The research addresses a gap in the limited availability of waste bank systems that integrate deposit management, e-wallet functionality, and e-commerce within a unified platform. By adopting the Agile Scrum approach, the system is designed to adapt to user needs through short development iterations and continuous evaluation. The testing results indicate that all system functionalities operate effectively, with a user acceptance rate (UAT) of 89.17%, categorized as Very Good. The novelty of this study lies in the direct integration between waste deposits and real economic exchange mechanisms, which has been shown to significantly enhance operational efficiency for various stakeholders while improving transparency in waste bank management
Application of Naïve Bayes Method for Student Performance Classification
Full text linkIn every school, students exhibit varying levels of performance, influenced by several factors such as parental support and involvement, participation in extracurricular activities, motivation levels, internet access for learning, teacher quality, peer influence, and learning difficulties. This study aims to classify student performance to identify those who may need additional support for improvement. The classification method employed in this research is the Naïve Bayes algorithm. The results indicate that the trained model successfully classified 25 out of 30 tested data points. The evaluation metrics achieved include a precision of 100%, recall of 80%, specificity of 100%, accuracy of 83%, and an F1-score of 89%
Performance Improvement of Periodic Reports with Materialized Views on Oracle Database System
Full text linkPeriodic reports in Management Information Systems (MIS) play a crucial role in organizational decision-making. However, rapid data growth can degrade query performance. This study aims to enhance the performance of periodic reports in the State Revenue Collection Module (SIM MPN) of the Ministry of Finance by utilizing Materialized Views (MVs) in the Oracle database system. A dataset comprising 294,503,898 rows from 2021 to 2023 was used as a sample. An analysis was conducted on eight types of periodic report queries that traditionally relied on standard views. These queries were then converted into MVs to reduce execution time. Testing was performed by comparing execution times between views and MVs across 24 queries executed in two different environments. The results showed that the average execution time using views was 2,047,417 ms, whereas with MVs, execution times were reduced to 41 ms in the first test, 17 ms in the second, and 12 ms in the third. These findings confirm that MVs significantly improve the performance of periodic reports by accelerating query execution. The practical implication of this study is the recommendation to implement MVs in systems with large data volumes to optimize report access speed. Future research can focus on optimizing MV refresh times and further analyzing factors affecting execution time under various usage scenarios
A Human-Centered Design Approach in Developing the “Jejak Cilik” Parenting App Prototype
Full text linkDevelopmental disorders in early childhood are a critical issue requiring serious attention. The role of parents in stimulating and monitoring child development is crucial; however, challenges remain, such as limited understanding and access to relevant information. To address this problem, this study developed a parenting app prototype called Jejak Cilik using a Human-Centered Design (HCD) approach. The development process followed four key phases: Discover, Define, Design, and Prototype & Test. The prototype was evaluated using heuristic evaluation with severity ratings to identify interface and user experience (UI/UX) issues. In addition, completion and duration metrics were assessed to evaluate the success rate and time required to complete assigned tasks. The heuristic evaluation, involving three evaluators, revealed three aspects that required immediate improvement and four aspects categorized as low-priority issues. Meanwhile, the completion and duration metrics, tested by 20 participants, showed promising results: all tasks were successfully completed by all users, with an average task completion time of less than one minute. However, during implementation, several issues were encountered that hindered task execution. Based on the findings from both evaluations, the prototype requires further refinement to enhance overall user experience
Security Analysis of the Final Project Information System (SITASI) Website using Penetration Testing Method
Full text linkThe Final Project Information System (SITASI) website plays a critical role in supporting academic administrative processes at the Faculty of Science and Technology, UIN Sultan Syarif Kasim Riau. This study aims to evaluate the website’s security level following recent maintenance using penetration testing, conducted with the OWASP Zed Attack Proxy (ZAP) tool. The testing revealed eight vulnerabilities, including two classified as medium risk, four as low risk, and two informational. The medium-risk issues involved the absence of an Anti-CSRF token and the lack of a Content Security Policy (CSP), both of which could expose the system to attacks such as CSRF and XSS. The low-risk findings included loading JavaScript from third-party domains, information disclosure via X-Powered-By and Server headers, and the absence of HTTP Strict Transport Security (HSTS). The two informational findings involved suspicious comments in the code and improper Cache-Control settings. Remediation actions were implemented based on OWASP security best practices, including the integration of CSRF tokens, configuration of CSP and HSTS headers, and removal of sensitive information from server responses. A follow-up evaluation confirmed that all identified risks had been successfully mitigated. This study highlights that penetration testing combined with standard-based mitigation is effective in enhancing web application security resilience, particularly within academic environments
Sentiment Analysis of MyBCA Application User Reviews using Naive Bayes, Random Forest, and Decision Tree
Full text linkIn today’s era of globalization, rapid technological advancements are driving innovation across various sectors, including the banking industry. One of the key digital innovations in banking is mobile banking (m-banking), which allows customers to perform transactions via smartphones. This study aims to analyze the sentiment of user reviews on the MyBCA application using three classification methods: Naive Bayes, Random Forest, and Decision Tree. A total of 5,000 user reviews were collected from the Google Play Store through web scraping techniques. The data was preprocessed using the TF-IDF weighting method and processed with Python programming language and the Scikit-Learn library. The dataset was split into 90% training data and 10% testing data. This study also applies the ISO 9126 standard for multi-label classification to assess software quality based on Usability, Efficiency, Functionality, Reliability, and Maintainability. Evaluation results indicate that Random Forest achieved the highest accuracy at 94.09%, outperforming Naive Bayes (81.77%) and Decision Tree (82.38%). This research contributes to the development of a sentiment-based evaluation method for mobile banking applications, integrating user feedback analysis with ISO 9126 quality standards, and offers a useful reference for improving digital banking services
Optimized Weight Evolutionary-based Support Vector Machine (SVM) Optimization for Comment Sentiment
Full text linkComments in questionnaire feedback carry sentiment meanings, such as positive, negative, or neutral. Each review comment on training services requires prompt and accurate follow-up to improve service quality. However, sentiment classification often demands significant time and effort to determine sentiments accurately. This study aims to enhance efficiency and accuracy in sentiment classification for training questionnaires. A comparative analysis was conducted using three algorithms: Naïve Bayes, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). The results indicate that SVM is the fastest and most accurate algorithm, with a training time of 3.067 seconds, 100 milliseconds faster than Naïve Bayes and 45.8 seconds faster than KNN. SVM achieved an accuracy of 60.81%, with an average sensitivity of 61%, specificity of 80%, and precision of 63%. Subsequently, this study integrated the Optimized Weight Evolutionary method to enhance SVM's accuracy and address attribute selection. Testing results showed a 2.16% improvement in SVM accuracy, bringing it to 63.10%. The training process was conducted on a dataset of 1,153 comments, with 90% of the data used for algorithm training. The combination of SVM and Optimized Weight Evolutionary proved effective in achieving more accurate sentiment classification. This study provides new insights into the application of sentiment classification, particularly for training feedback. Optimizing the algorithm can help training companies respond more effectively to comments and improve overall service quality
Implementation of Fuzzy Multiple Attribute Decision Making (FMADM) and Simple Additive Weighting (SAW) for Selecting the Best Stocks
Full text linkStock investment is an attractive option for modern society to enhance long-term asset value, despite significant risks, especially for novice investors who often face limited understanding of fundamental stock analysis. Such analysis involves various complex financial indicators. This study combines the Fuzzy Multiple Attribute Decision Making (FMADM) method and the Simple Additive Weighting (SAW) method to assist investors in selecting the best stocks in the banking sector. FMADM is applied to address data uncertainty using a fuzzy approach, while SAW calculates the final score based on criteria weights and performance. The research data were obtained from company financial reports for the 2019–2023 period, focusing on seven key criteria: Return on Assets (ROA), Return on Equity (ROE), Earnings Per Share (EPS), Net Profit Margin (NPM), Price-to-Book Value (PBV), Debt-to-Equity Ratio (DER), and Dividend Yield (DY). The study aims to develop a decision support system to simplify the investment analysis process while reducing the risk of decision-making errors for investors. The findings indicate that alternative A35/Bank BTPN Syariah (BTPS) ranked first with a final score of 0.7120, followed by A42/Bank Mega (MEGA) in second place with a score of 0.7074, and A08/Bank Central Asia (BBCA) in third place with a score of 0.6988. This system provides a practical solution for more structured and efficient investment decisions
Optimizing Feature Selection in Sentiment Analysis of Bank Saqu: A Comparative Study of SVM and Random Forest using Information Gain and Chi-Square
Full text linkThe selection of an optimal feature selection method is a crucial factor in improving the accuracy and efficiency of text classification models. Irrelevant features can degrade model performance, increase computational complexity, and lead to overfitting. Although various feature selection techniques have been employed in sentiment analysis, systematic studies comparing the effectiveness of Information Gain and Chi-Square in enhancing classification performance remain limited. This study aims to evaluate and optimize the impact of different feature selection methods on the performance of Support Vector Machine (SVM) and Random Forest (RF) in sentiment analysis. Experiments were conducted using eight testing schemes, including models without feature selection, with Information Gain, Chi-Square, and a combination of both. The results showed that SVM with Chi-Square achieved the highest accuracy at 93%, while Random Forest with Chi-Square achieved the best performance at 91%. These findings indicate that Chi-Square is more effective than Information Gain in improving accuracy, and that SVM outperforms Random Forest in text classification tasks. In conclusion, selecting the appropriate feature selection method significantly contributes to enhancing the accuracy of text classification models. This research can serve as a reference for optimizing feature selection techniques in the development of more accurate and efficient machine learning-based systems