Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Jurnal Riset dan Aplikasi Mahasiswa Informatika (JRAMI)
Muhammad Fabian Hartono
Universitas Trilogi

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Pengembangan Hybrid Recommender System Menggunakan Scikit-Learn dan Pandas untuk Rekomendasi Film Muhammad Fabian Hartono; Opitasari Opitasari; Ivan Firdaus
Jurnal Riset dan Aplikasi Mahasiswa Informatika (JRAMI) Vol. 7 No. 02 (2026): Jurnal Riset dan Aplikasi Mahasiswa Informatika (JRAMI)
Publisher : Program Studi Teknik Informatika, FTIK, Universitas Indraprasta PGRI

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30998/jrami.v7i02.1190

Abstract

The rapid growth of digital streaming services has led to an increase in the amount of available film content, creating a problem of information overload for users when selecting content that matches their preferences. Therefore, recommendation systems have become an essential solution to help users efficiently and personally discover relevant films. This study aims to develop a film recommendation system based on a hybrid recommender system by combining collaborative filtering and content-based filtering methods to improve the accuracy and diversity of recommendations. The CF method is used to leverage interaction patterns and similarities in preferences among users, while the CBF method utilizes film content features such as genre to determine item similarity. The system was implemented using the Python programming language with the scikit-learn and pandas libraries for data processing and model development. The dataset used is MovieLens 100k, consisting of 100,000 ratings from 943 users for 1,682 movies, along with movie metadata. System performance was evaluated using the mean squared error and precision@K metrics. The test results indicate that the hybrid system achieved a precision@5 value of 0.6000, indicating that 60% of the recommendations provided were relevant to user preferences. Furthermore, the hybrid approach proved to be more stable and accurate than single methods and was able to address cold-start problems and data sparsity. Thus, the developed system is capable of providing more relevant, diverse, and personalized movie recommendations. This study demonstrates that the hybrid approach is an effective solution for improving the quality of recommendation systems, particularly in user-data-based movie recommendation applications.  
Co-Authors Ivan Firdaus Opitasari Opitasari