Research Article
Big Data-Enabled Decision Support Using a Hybrid Design Science and Machine Learning Model
@INPROCEEDINGS{10.4108/eai.5-12-2025.2363851, author={Abdul Hamid K and Mansur AS and Sugiharto Sugiharto and Sitti Subaedah and Andi Bahar}, title={Big Data-Enabled Decision Support Using a Hybrid Design Science and Machine Learning Model}, proceedings={Proceedings of the 10th Annual International Seminar on Transformative Education and Educational Leadership, AISTEEL 2025, 5 December 2025, Medan, North Sumatera Province, Indonesia}, publisher={EAI}, proceedings_a={AISTEEL}, year={2026}, month={7}, keywords={Big Data; Decision Support System; Machine Learning; Higher Education}, doi={10.4108/eai.5-12-2025.2363851} }- Abdul Hamid K
Mansur AS
Sugiharto Sugiharto
Sitti Subaedah
Andi Bahar
Year: 2026
Big Data-Enabled Decision Support Using a Hybrid Design Science and Machine Learning Model
AISTEEL
EAI
DOI: 10.4108/eai.5-12-2025.2363851
Abstract
This study evaluated the effectiveness and feasibility of an AI-based e-learning system using quantitative measures and expert validation aligned with the ADDIE evaluation phase. System performance was assessed through percentage improvement analysis, user satisfaction indices derived from questionnaires, and expert validation conducted via Focus Group Discussion (FGD). Percentage scores were interpreted using a five-level classification scale. The results show notable improvements in learning efficiency (58%), feedback response time (65%), interactivity (54%), and student motivation (61%). User satisfaction reached 89%, indicating strong acceptance and usability. Expert evaluation further confirmed system readiness, with feasibility scoring 92% and pedagogical relevance 85%. These findings demonstrate that integrating machine learning–based adaptive recommendations, automated feedback, learning analytics, and NLP-based chatbot support can effectively enhance learning performance, engagement, and instructional responsiveness. Overall, the proposed AI-based e-learning system is a feasible and pedagogically relevant solution for supporting adaptive and data-driven learning in graduate education.
