COMPARISON MODELS OF MACHINE LEARNING FOR MOVIE RECOMMENDATION SYSTEMS
Ключевые слова:
recommendation system, analysis of machine learning approaches, Surprise library, collaborative filtering.Аннотация
The trend of the Internet makes the presentation of the right content for the right user inevitable. To
this end, recommendation systems are used in areas such as music, books, movies, travel planning, e-commerce,
education, and more. One of the most popular recommendation systems in the world is Netflix, which generated
record profits during quarantine in the first quartile of 2020. The systematic approach of recommendations is based
on the history of user selections, likes and reviews, each of which is interpreted to predict future user selections. This
article provides a meaningful analysis of various recommendation systems, such as content-based, collaborative
filtering and popularity. We reviewed 7 articles published from 2005 to 2019 to discuss issues related to existing
models. The purpose of this article is to compare machine learning algorithms in the Surprise library for a
recommendation system. Recommendation system has been implemented and quality has been evaluated using the
MAE and RMSE metrics.
