Sistem Rekomendasi TV Series Berdasarkan Genre Menggunakan Algoritma KNN

Deni Fahrizal
Abdul Halim Hasugian

Jurnal Sains dan Teknologi
Vol. 4 No. 4 (2025): Agustus 2025

The problem of choice overload on TV series streaming platforms often makes it difficult for users to find content that suits their preferences. To address this challenge, this study develops a Content-Based Filtering-based recommendation system by applying the K-Nearest Neighbor (KNN) algorithm and the Jaccard Similarity metric. The designed system analyzes users' genre preferences, such as Drama, Sci-Fi, and Comedy, while integrating rating, popularity, and release year factors to generate more personalized recommendations. Evaluation of 500 TV series titles from the TMDB API shows a high level of accuracy, with Precision and Recall reaching 1.0 for specific genre preferences, as well as stable performance with an F1-Score of 0.67 for cross-genre preferences. These findings prove that the proposed model is effective in reducing choice overload and significantly improving the user experience in exploring content on streaming platforms. Furthermore, this approach has the potential to be further developed by integrating sentiment analysis and real-time audience behavior data to generate increasingly adaptive and relevant recommendations.