Scholarship Recommendation System Using Content-Based Filtering Approach

Abstract

The numerous prospective scholarship recipients who face difficulties in finding information that matches their educational profiles form the background of this research. The objective of this study is to design and develop a web-based scholarship recommendation system using the Content-Based Filtering (CBF) method and to determine the accuracy level of the Euclidean Distance algorithm through precision testing. This research is a descriptive qualitative study utilizing the Rapid Application Development (RAD) approach, which includes Requirement Planning, User Design, Construction, and Cutover phases. Data sources consist of scholarship information from official institutions and user profile inputs. Data collection was conducted through literature studies, observations, and questionnaires. Functionality testing was performed using Black Box Testing, while recommendation accuracy analysis utilized the Precision method based on Euclidean Distance. The results showed that a web-based system using MERN Stack technology was successfully developed and capable of presenting a list of scholarships relevant to user profile metadata. Based on the testing, this algorithm achieved an average accuracy of 92%, proving that the system works optimally in prioritizing relevant scholarships and effectively overcomes the cold-start problem. In conclusion, this system is mathematically validated as a decision support tool for scholarship seekers. Suggestions for future development include integrating the system with external APIs such as PDDikti for automatic data validation and improving the text-matching mechanism to avoid rigid exact matches.