Implementasi Sistem Rekomendasi Film Berbasis Collaborative Filtering: Studi Kasus MovieLens

Syuyukh, Fakhrusy
Wicaksono, Galih Wasis
Nuryasin, Ilyas

Jurasik (Jurnal Riset Sistem Informasi dan Teknik Informatika)
Vol 10, No 2 (2025): Edisi Agustus

Movie recommendation systems often face challenges such as data sparsity and cold-start. This study compares three Collaborative Filtering (CF) approaches: User-Based, Item-Based, and Singular Value Decomposition (SVD) on the MovieLens dataset. Evaluation is performed using RMSE, MAE, Precision@10, and Recall@10 metrics. The results show that SVD with n_factors = 50 provides the best accuracy (RMSE 0.8695; MAE 0.6682), while Item-Based with k = 10 excels in recommendation relevance (Precision@10 0.5510; Recall@10 0.6267). This study provides a reference in selecting a recommendation algorithm based on the context of system needs.