Article Per Year (5 Year)
p-Index From 2020 - 2025
0.751
P-Index
This Author published in this journals
All Journal IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Sistemasi: Jurnal Sistem Informasi Journal of Information Systems and Informatics
Purba, Susi Eva Maria
Unknown Affiliation
Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
A Mamdani FIS to Monitor Programmer Performance on GitHub Purba, Susi Eva Maria; Wardoyo, Retantyo
IJCCS (Indonesian Journal of Computing and Cybernetics Systems) Vol 18, No 2 (2024): April
Publisher : IndoCEISS in colaboration with Universitas Gadjah Mada, Indonesia.

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22146/ijccs.88575

Abstract

A collaborative activity used to accomplish shared objectives is teamwork. It is essential to know how unequal contributions can inhibit team members' chances to give their all in achieving these objectives. It will be necessary to manage resources in this joint approach. Monitoring each team member’s performance in one technique to do this. In previous research, performance measurement was designed using Prometer with several parameters, utilizing the crisp set at each stage. This study developed the method by adding variables and utilizing fuzzy logic, which can consider the membership value for each value involved. The membership value considered for each variable is expected to provide a significant assessment of each team working on developing software projects using the GitHub platform. The results will be monitored based on the involvement of each collaborator in project work through the data recorded in the pull requests, issues, commits, additions code, and deletion code variables. The results obtained by utilizing the variables and several rules that have been designed with the Mamdani implication function are then compared with the observations obtained by the Project Manager so that an accuracy value of 86.67% is accepted for the use of inclusive and exclusive rules (operand AND).
A Comparative Study of Drug Prediction Models using KNN, SVM, and Random Forest Purba, Susi Eva Maria
Journal of Information System and Informatics Vol 7 No 1 (2025): March
Publisher : Universitas Bina Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalisi.v7i1.1013

Abstract

Accurate drug classification is essential in medical decision-making to ensure patients receive appropriate prescriptions based on their physiological and biochemical characteristics. This study compares the performance of K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Random Forest models in predicting drug prescriptions using patient attributes such as age, sex, blood pressure, cholesterol level, and sodium-to-potassium ratio. The dataset, obtained from Kaggle, was preprocessed and split into training and testing sets to evaluate model performance using accuracy as the primary metric. The results indicate that Random Forest outperformed KNN and SVM, achieving a perfect test accuracy of 100%, demonstrating superior generalization and robustness. SVM also performed well, with a test accuracy of 97.50%, while KNN achieved the lowest accuracy of 70%, indicating its limitations in handling complex feature interactions. These findings highlight the effectiveness of ensemble learning methods in medical classification tasks, suggesting that Random Forest is the most suitable model for drug prediction. Furthermore, the potential applications of these findings in clinical settings could enhance treatment outcomes and patient care. Future research should explore feature engineering techniques, larger datasets, and additional machine learning approaches to enhance predictive accuracy and applicability in real-world healthcare settings.
Enhancing Hate Speech Detection: Leveraging Emoji Preprocessing with BI-LSTM Model Amalia, Junita; Tambunan, Sarah Rosdiana; Purba, Susi Eva Maria; Simanjuntak, Walker Valentinus
Journal of Information System and Informatics Vol 7 No 2 (2025): June
Publisher : Universitas Bina Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalisi.v7i2.1147

Abstract

Microblogging platforms like Twitter enable users to rapidly share opinions, information, and viewpoints. However, the vast volume of daily user-generated content poses challenges in ensuring the platform remains safe and inclusive. One key concern is the prevalence of hate speech, which must be addressed to foster a respectful and open environment. This study explores the effectiveness of the Emoji Description Method (EMJ DESC), which enhances tweet classification by converting emojis into descriptive text or sentences. These descriptions are then encoded into numerical vector matrices that capture the meaning and emotional tone of each emoji. Integrated into a basic text classification model, these vectors help improve detection performance. The research examines how different emoji preprocessing strategies affect the performance of a BI-LSTM model for hate speech classification. Results show that removing emojis significantly reduces accuracy (68%) and weakens the model’s ability to distinguish between hate and non-hate speech, due to the loss of valuable semantic context. In contrast, retaining emoji semantics either through textual descriptions or embeddings boosts classification accuracy to 93% and 94%, respectively. The highest performance is achieved through emoji embedding, highlighting its ability to capture subtle non-verbal cues critically for accurate hate speech detection. Overall, the findings emphasize the importance of incorporating emoji-aware preprocessing techniques to enhance the effectiveness of social media content classification.
Integrating Agile and Business Metrics into Backlog Prioritization: A Case Study at PT. XYZ Purba, Susi Eva Maria; Tambunan, Katrina Arlyanti
Sistemasi: Jurnal Sistem Informasi Vol 14, No 6 (2025): Sistemasi: Jurnal Sistem Informasi
Publisher : Program Studi Sistem Informasi Fakultas Teknik dan Ilmu Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32520/stmsi.v14i6.5244

Abstract

Backlog prioritization is an essential component of Agile development because it makes sure that resources are used in the best way possible and that business value is maximized. The Effort-influence Matrix gives you a way to prioritize items in your backlog based on how much effort they will take and how much influence they might have. However, just prioritizing them doesn't mean you'll be successful in the long term. Integrating Agile Metrics—such as velocity, cycle time, and lead time—with Business Metrics—such as customer satisfaction, retention, and market adoption—offers a more comprehensive approach to guiding decision-making. This study examines how Product Owners at PT. XYZ applies the Effort-Impact Matrix while incorporating Agile and Business Metrics to align development priorities with organizational objectives. This study employed qualitative research design, drawing on structured interviews, project documentation, and literature review. The findings show that combining prioritizing frameworks with performance indicators improves decision-making, increases alignment with company goals, and leads to more predictable delivery outcomes. This study contributes to the literature by being among the few to empirically demonstrate how Agile and Business Metrics can be systematically integrated into backlog prioritization using the Effort-Impact Matrix.
Co-Authors JUNITA AMALIA Retantyo Wardoyo Sarah Rosdiana Tambunan Simanjuntak, Walker Valentinus Tambunan, Katrina Arlyanti