Implementation of a Decision Support System Based on the Simple Additive Weighting (SAW) Method and Apriori Algorithm for Student Achievement Evaluation

Figure Ilustration AI

FORMOSA NEWS - Jakarta - AI-Powered System Boosts Student Achievement Evaluation Accuracy to 87 Percent. Researchers at Universitas Budi Luhur have successfully developed an artificial intelligence-driven Decision Support System (DSS) that evaluates student achievement with 87 percent accuracy, eliminating human bias and manual processing errors in academic recognition. Developed by Mochammad Bagus Priyantono and Deni Mahdiana and published in 2026, the system combines multi-criteria mathematical ranking with machine learning pattern recognition. The breakthrough offers educational institutions an automated, objective, and transparent mechanism to identify high-achieving students based on complex multi-dimensional data.

Selecting outstanding students has long presented a challenge for school administrators and educational policymakers. Conventional evaluation processes rely heavily on manual calculations across diverse indicators, including academic performance, attendance, discipline, extracurricular achievements, and classroom participation. Without a standardized computational framework, manual evaluations often introduce subjectivity, risk human error, and suffer from a lack of transparency, which can cause dissatisfaction among students and faculty.

To address these operational inefficiencies, the research team engineered a dual-method decision-making framework. The system uses the Simple Additive Weighting (SAW) method to compute structured student rankings alongside the Apriori algorithm to discover behavioral pattern associations and validate data consistency.

The methodology integrates two complementary computational approaches into a unified software environment:

• Simple Additive Weighting (SAW) Ranking: The SAW algorithm assigns mathematical weights to five core performance metrics based on institutional priorities. Academic score carries the highest weight at 30 percent, followed by attendance at 20 percent, competition achievements at 20 percent, discipline at 15 percent, and classroom activeness at 15 percent. The system normalizes these raw metrics into uniform scales and computes a final preference score for each student to establish an objective ranking.
  
• Apriori Algorithm Validation: Once rankings are generated, the Apriori data-mining algorithm converts student profile metrics into transaction datasets. By setting minimum support and confidence thresholds across historical student data, Apriori extracts frequent attribute patterns—such as the correlation between high attendance, discipline, and top-tier academic performance. The system checks new student evaluations against these established historical rules, labeling data combinations as valid or invalid to ensure analytical consistency.
  

The research team evaluated the dual-method system using a dataset of 100 student academic records, allocating 85 records as historical training data and 15 records as experimental testing data. When compared directly against manual evaluation results produced by an expert teaching committee, both the Simple Additive Weighting ranking mechanism and the Apriori pattern validation process achieved an accuracy rate of 87 percent.

The findings demonstrate significant real-world benefits for school administration, educational technology deployment, and public policy. By automating multi-criteria data processing, institutions can reduce administrative overhead, shorten evaluation cycles, and ensure complete transparency in awarding academic honors or merit scholarships. Furthermore, the association rules identified by the Apriori algorithm provide administrators with data-driven insights into key behavioral indicators that drive student success, allowing for earlier targeted interventions for struggling students.

Author Profiles
Mochammad Bagus Priyantono: Scholar and researcher affiliated with the Department of Computer Science and Information Technology at Universitas Budi Luhur, specializing in Decision Support Systems, Multi-Criteria Decision Making (MCDM), and Data Mining algorithms.
Deni Mahdiana: Senior faculty member and researcher at Universitas Budi Luhur, possessing extensive academic expertise in Information Systems engineering, computational decision models, and software architecture.

Source
Mochammad Bagus Priyantono, Deni Mahdiana. Implementation of a Decision Support System Based on the Simple Additive Weighting (SAW) Method and Apriori Algorithm for Student Achievement Evaluation. Formosa Journal of Computer and Information Science (FJCIS). Vol. 5, No. 2, Hal. 279-294).
DOI : https://doi.org/10.55927/fjcis.v5i2.16835
URL: https://journal.formosapublisher.org/index.php/fjcis