Automatic movie ratings prediction using machine learning
Mladen Marović, Marko Mihokovic, Mladen Mikša, Sinisa Pribil, Alan Tus · International Convention on Information and Communication Technology, Electronics and Microelectronics · 2011
Recommendation systems that model users and their interests are often used to improve various user services. Such systems are usually based on automatic prediction of user ratings of the items provided by the service. This paper presents an overview of some of the methods for automatic ratings prediction in the domain of movie ratings. The chosen methods are based on various approaches described in related papers. During the prediction process both the user and item features can be used. For the purpose of this paper, data was gathered from the publicly available movie database IMDb. The paper encompasses the implementation of the chosen methods and their evaluation using the gathered data. The results show an improvement in comparison to the chosen baseline methods.