Samuel Situmeang

Samuel Situmeang

M.Sc. in Computer Science and Information Engineering and Lecturer at Del Institute of Technology

Publications

A Cluster and Association Analysis Visualization using Moodle Activity Log Data

Published in The International Journal of Informatics and Communication Technology (IJ-ICT), 2022

The course activity log is where a learning management system (LMS) like Moodle keeps track of the various learning activities. In order to conduct a quicker and more in-depth examination of the students' behaviors, the instructor may either directly examine the log or make use of more complex methodologies such as data mining. The majority of the proposed methods for analyzing this log data center mostly on predictive analysis. In this research, cluster analysis and association analysis, two separate data mining functions, are investigated in order to analyze the log. The students' activities are used in the cluster analysis performed with K-Means++, and the association analysis performed with Apriori is used to investigate the connections between the students' various activities. A dashboard presentation of the findings is provided in order to facilitate clearer comprehension. Based on the findings of the analysis, it can be concluded that the structure of the student cluster is medium, whereas the association between the activities undertaken by students is positively correlated and well-balanced. The subjective review of the dashboard reveals that the visualization is already sufficient, but there are some recommendations for making it even better.

Recommended citation: Reimondo Tamba, A., Lumbantoruan, K., Pakpahan, A., & Situmeang, S. (2023). A Cluster and Association Analysis Visualization using Moodle Activity Log Data. International Journal of Informatics and Communication Technology (IJ-ICT), 12(2), 150. https://doi.org/10.11591/ijict.v12i2.pp150-161 https://doi.org/10.11591/ijict.v12i2.pp150-161

Android Malware Classification Based on Permission Categories Using Extreme Gradient Boosting

Published in Proceedings of the 5th International Conference on Sustainable Information Engineering and Technology, 2020

Mobile malware has become the centerpiece of most security and privacy threats on the Internet. Especially with the openness of the Android market, many malicious apps are hiding in a large number of applications, which makes malware detection more challenging. In this study, eXtreme Gradient Boosting (XGBoost) is used to establish the Android-based malware detection and classification framework. The framework utilizes APK permission categories extracted from Android applications. The comparison of modeling results demonstrates that the XGBoost is especially suitable for Android malware classification and can achieve 74.40% of F1-score with real-world Android application sets.

Recommended citation: T. N. Turnip, A. Situmorang, A. Lumbantobing, J. Marpaung, and S. I. G. Situmeang, "Android Malware Classification Based on Permission Categories Using Extreme Gradient Boosting," in ACM International Conference Proceeding Series, Nov. 2020, pp. 190-194. doi: 10.1145/3427423.3427427. https://dl.acm.org/doi/pdf/10.1145/3427423.3427427

Movie Summarization based on Indonesian Subtitles with Restricted Boltzmann Machine

Published in Proceedings of the 2019 International Conference on Sustainable Information Engineering and Technology, SIET 2019, 2019

A summary of a movie provides a concise description of the movie both concretely and abstractly. Automatic movie summarization can be applied to get a movie summary. In this study, we propose a movie summarization method on Indonesian subtitles with five stages, namely: preprocessing, feature extraction, feature enhancement, sentence ranking, and summary generation. The preprocessing consist of document segmentation, paragraph segmentation, case folding, normalization, stopword removal, and part-of-speech tagging. Twelve intrinsic features, namely: number of thematic words, sentence position, sentence length, sentence position relative to paragraph, number of proper noun, number of numerals, number of verbs, number of noun, term frequency-inverse sentence frequency, sentence to centroid similarity, bi-gram key phrase list, and tri-gram key phrase list are extracted and enhanced with Restricted Boltzmann Machine (RBM) to improve the quality of the summary. Results indicate that the summary produced is understandable and has the name of the main character in the provided text. Compare to movie summarization without RBM, RBM improves the average quality of the summary of 6% (F1-score). Questionnaire results also show that the movie summarization developed in this study produces a summary that is suitable for the provided subtitles.

Recommended citation: S. I. G. Situmeang, R. K. Lubis, F. J. N. Siregar, and B. J. D. C. Panjaitan, "Movie Summarization based on Indonesian Subtitles with Restricted Boltzmann Machine," in Proceedings of 2019 4th International Conference on Sustainable Information Engineering and Technology, Sep. 2019, pp. 338-342. doi: 10.1109/SIET48054.2019.8986127. https://ieeexplore.ieee.org/abstract/document/8986127

Badminton Video Analysis based on Spatiotemporal and Stroke Features

Published in Proceedings of the 2017 ACM on International Conference on Multimedia Retrieval, 2017

Most of the broadcasted sports events nowadays present game statistics to the viewers which can be used to design the gameplay strategy, improve player's performance, or improve accessing the point of interest of a sport game. However, few studies have been proposed for broadcasted badminton videos. In this paper, we integrate several visual analysis techniques to detect the court, detect players, classify strokes, and classify the player's strategy. Based on visual analysis, we can get some insights about the common strategy of a certain player. We evaluate performance of stroke classification, strategy classification, and show game statistics based on classification results.

Recommended citation: W. T. Chu and S. Situmeang, "Badminton Video Analysis based on Spatiotemporal and Stroke Features," in ICMR 2017 - Proceedings of the 2017 ACM International Conference on Multimedia Retrieval, 2017, pp. 448-451. doi: 10.1145/3078971.3079032. https://dl.acm.org/doi/pdf/10.1145/3078971.3079032