IJID (International Journal on Informatics for Development)

K-Means Clustering of Social Studies Performance at Junior High School

Tundo — 2024-12-29

This study aims to optimize the use of technology in evaluating student performance by grouping students based on their abilities. The main issues include the underutilization of technology, the absence of an appropriate evaluation system for different levels of student ability, and ineffective methods for grouping students. The K-Means Clustering algorithm was chosen because it has proven effective in grouping academic data in various studies. The data used includes Daily Knowledge Scores (DKS), Daily skill scores (DSS), Mid-term Summative Scores (MSS), End-of-Year Summative Scores (ESS), and Grade Report (GR). The data was analyzed using the CRISP-DM methodology with the help of RapidMiner. The results showed that 28.63% of students were classified as having excellent performance, 50.21% as having good performance, and 21.16% as having moderate performance. The Davies-Bouldin Index score of 1.713 for K=3 was considered sufficient for distinguishing the different student performance groups. The results of this study are expected to help schools provide learning support that better aligns with student needs. Future research is recommended to focus on optimizing the number of clusters (K), applying this method to other subjects, and integrating it with e-learning platforms for real-time student performance monitoring.

Analyzing Customer Loyalty Levels through Segmentation in Aesthetic Clinics Using K-Means and RFAM

Sinarring Azi Laga — 2024-12-29

Effective customer segmentation is crucial in optimizing marketing strategies, particularly in customer-oriented aesthetic clinics. This research aims to enhance customer segmentation in aesthetic clinics using a K-Means approach based on the RFAM (Recency, Frequency, Average-Monetary) model. This approach is utilized to leverage historical customer data to identify customer segments based on their purchasing behavior, including visit frequency, average purchase amount, and the last time they visited the clinic. The K-Means clustering method maps customers into homogeneous groups, enabling aesthetic clinics to adapt more focused and personalized marketing strategies. The research results indicate insights obtained from the analysis and interpretation of RFAM conducted on 493 data points, resulting in the formation of two distinct clusters. In Cluster 1, denoting low loyalty, there are 156 customers, while Cluster 2 comprises 337 customers, reflecting high loyalty. Practical implications of this research include improvements in service customization and promotions tailored to customer needs and preferences. In conclusion, the K-Means approach based on the RFAM model can be utilized as an effective tool to enhance customer segmentation in the aesthetic clinic industry.

Leveraging Ontology-Driven Machine Learning for Public Policy Analysis: A Systematic Review of Social Media Applications

ADMAS Kero — 2024-12-29

As social media platforms increasingly serve, machine learning techniques are formulated with particular ontologies, which furnish invaluable resources. This qualitative literature review investigates the incorporation of ontology-driven machine learning methodologies for analysing public policy utilizing social media data. This review encompasses findings from scholarly research published between 2019 and 2024 that apply ontologies to enhance models' interpretation, precision, and flexibility across diverse sectors, including health, environment, economy, and culture. An integrated methodology is adopted to identify, select, and evaluate pertinent studies by scrutinizing elements such as genre ontology, machine learning, existing literature, and evaluation metrics. The findings indicate that the ontology-centric framework facilitates the extraction process and semantic analysis, ultimately contributing to a more nuanced comprehension of unstructured data. Nonetheless, obstacles persist in ontology development concerning capacity enhancement, data integrity, and ethical considerations. The review concludes with a discourse on the ramifications for policymakers and researchers who may leverage these insights to guide decision-making, and scholars are now urged to confront limitations and investigate novel platforms, metrics, and ethical frameworks. The review underscores the potential of ontology-driven machine learning as a formidable strategy in the advancement of policy research and social analysis.

Assessing AI Integration in Islamic Higher Education: A Mixed-Methods Fishbone Diagram Analysis

Aan Ansori — 2025-01-22

The integration of Artificial Intelligence in higher education has shown significant potential to improve the efficiency and effectiveness of learning. The strategic implementation of AI in Indonesian State Islamic Higher Education Institutions fosters innovative pedagogy and improved academic performance. This study employs the Fishbone Diagram approach to systematically analyze Artificial Intelligence's impact on Indonesian State Islamic Higher Education Institutions education, identifying key factors influencing implementation. The method employs a reverse-cause analysis, mapping factors contributing to a primary issue, and identifying underlying causes and sub-factors. Findings highlight the crucial roles of technological infrastructure, human resource readiness, supportive policies, adaptive curriculum design, and organizational culture. This study underscores the necessity of integrated AI adoption frameworks in Indonesian Islamic higher education, harmonizing technological advancement with Islamic pedagogical principles. This study offers a foundational framework guiding Indonesian State Islamic Higher Education Institutions in developing sustainable and ethical AI policies. Comprehensive AI policies and strategies are essential for PTKIN to harmonize innovation with Islamic principles.

Improving Osteosarcoma Detection through SMOTE-Driven Machine Learning Approaches

Muhammad Ainul Fikri — 2025-02-07

Osteosarcoma is an aggressive and highly malignant bone cancer primarily affecting adolescents and young adults, with males being more commonly affected. Although deep learning models such as YOLO (95.73% accuracy) and VGG19 (95.25% accuracy), have demonstrated effectiveness in osteosarcoma detection, their large model sizes and extensive computational requirements limit their feasibility in resource-constrained environments. This study proposes a lightweight AI approach that optimizes osteosarcoma detection while maintaining high diagnostic accuracy, leveraging machine learning models under 5MB, manually or semi-automatically extracted features, and SMOTE for data balancing. Experimental results show that Random Forest, SVM, and XGBoost achieve accuracies of 94.70%, 94.23%, and 94.39%, respectively, closely matching the performance of YOLO and VGG19 while maintaining computational efficiency. Furthermore, the inference time for SVM is under one second (0.97s), demonstrating the speed advantage of lightweight models. These findings highlight the potential of small-size (lightweight) machine learning models to deliver high diagnostic accuracy with minimal computational requirements, providing a scalable and practical solution for early osteosarcoma detection in resource-limited settings. By balancing simplicity, efficiency, and high performance, this study establishes a new benchmark for achieving state-of-the-art results with lightweight models and paving the way for improved healthcare accessibility in underserved regions.