Development and Quality Analysis of Decision Support Systems as Software for Scholarship Recommendation in Higher Education

Abstract

This study aimed to develop a system that provides scholarship data, recommendations based on students' ability and that ensures that the developed system provides appropriate recommendations according to the students' ability by examining the results of system recommendations on academic, organizational, and student achievement. The research design used a Research and Development (R & D) method and procedures in the development of a scholarship recommendation system using the Extreme Programming (XP) development procedure with four stages: planning, design, coding, and testing. Testing of recommendations was performed by using a guide confusion matrix which tests seven aspects such as accuracy, recall, precision, prevalence, error rate, false positive rate, and specificity. The results of the research showed that: (1) the recommendation system for the selection of scholarships in universities resulted in recommendations using a content-based recommendation method with the weighting of the term frequency-inverse document frequency. (2) Testing the accuracy of the recommendation resulted in a 100% success rate.