Penerapan Content-Based Filtering untuk Rekomendasi Film Netflix Berdasarkan Genre dan Klasifikasi Usia Penonton

Abstract

Recommendation systems have become an essential part of improving user experience, particularly in movie selection. With the ever-increasing number of movie titles, users often experience difficulty in determining which movies suit their interests and needs. Mismatches in movie selection, particularly in terms of age classification, can have negative impacts, particularly for children exposed to inappropriate content. Therefore, a recommendation system is needed that not only considers user preferences but also takes into account age restrictions. This study aims to build a movie recommendation system using a content-based filtering approach, considering two main aspects: genre and age classification. In its implementation, the Term Frequency-Inverse Document Frequency (TF-IDF) method is used as a word weighting technique for movie content information. This weighting is then used to calculate the level of similarity between movies using the Cosine Similarity method. This system is designed to be able to recommend relevant movies based on user preference input, such as the desired movie genre and the appropriate age classification. An evaluation was conducted to measure the extent to which the system is able to provide recommendations that match user preferences. The evaluation results show a similarity value of 1.00 for the genre aspect and 1.00 for the age classification aspect. This score indicates that the system successfully recommended highly relevant films that met the user's specified criteria. Therefore, the developed recommendation system effectively filters and suggests films that are not only content-appropriate but also safe for viewing based on the specified age restrictions.