Analisis dan implementasi Content-Based Filtering dengan Cosine Similarity untuk sistem rekomendasi tugas akhir mahasiswa

Abstrak

This study aims to analyze and implement a student thesis recommendation system based on the Content-Based Filtering approach, using the Cosine Similarity method. The proposed system is designed to assist students in identifying relevant thesis topics based on their interests and prior research. The methods used in this research include data collection from a repository through the OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting), text preprocessing by removing Indonesian stop words, the formation of TF-IDF matrices, and the calculation of document similarity using cosine similarity. The experimental results show that the system successfully provides relevant recommendations with satisfactory precision and recall, as well as fast response times, making it suitable for real-world applications. These findings suggest that the Content-Based Filtering approach, combined with cosine similarity, is effective at capturing content similarity in student theses and providing recommendations that align with user needs. Therefore, the developed system can serve as an efficient and accurate solution to support students in selecting appropriate thesis topics and also facilitate access to relevant academic references.