Detail Cantuman

Perkembangan industri hiburan, khususnya drama Korea, menghadirkan tantangan information overload bagi pengguna karena banyaknya pilihan yang tersedia di berbagai platform digital. Untuk mengatasi hal tersebut, penelitian ini mengembangkan sistem rekomendasi drama Korea menggunakan metode Item-Based Collaborative Filtering (IBCF) dengan pendekatan Pearson Correlation Coefficient (PCC). Data yang digunakan berasal dari platform Kaggle, dengan atribut utama berupa user_id, title, dan overall_score. Tahapan penelitian meliputi akuisisi data, data cleaning, transformasi data, pembentukan matriks user–item, perhitungan kesamaan antar item menggunakan PCC, serta implementasi fungsi rekomendasi berbasis weighted sum. Proses evaluasi dilakukan dengan membagi data ke dalam training set dan testing set, kemudian mengukur akurasi prediksi menggunakan metrik Mean Absolute Error (MAE) dan Root Mean Square Error (RMSE). Hasil penelitian menunjukkan bahwa metode IBCF dengan PCC mampu menghasilkan rekomendasi yang relevan, meskipun tetap dipengaruhi oleh kendala sparsity dan cold-start. Dengan demikian, sistem yang dibangun dapat membantu pengguna menemukan drama baru sesuai preferensi, sekaligus memberikan kontribusi dalam pengembangan sistem rekomendasi hiburan yang lebih adaptif.

The development of the entertainment industry, particularly Korean dramas, presents users with the challenge of information overload due to the large number of choices available on various digital platforms. To overcome this, this study developed a Korean drama recommendation system using the Item-Based Collaborative Filtering (IBCF) method with the Pearson Correlation Coefficient (PCC) approach. The data used came from the Kaggle platform, with the main attributes being user_id, title, and overall_score. The research stages included data acquisition, data cleaning, data transformation, user–item matrix formation, calculation of item similarity using PCC, and implementation of a weighted sum-based recommendation function. The evaluation process was carried out by dividing the data into training and testing sets, then measuring the prediction accuracy using the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) metrics. The results show that the IBCF method with PCC is capable of generating relevant recommendations, although it is still affected by sparsity and cold-start constraints. Thus, the developed system can help users find new dramas according to their preferences, while contributing to the development of a more adaptive entertainment recommendation system.