Using Filters in Time-based Movie Recommender Systems

摘要

A movie recommendation system is one which uses data about the user, data about the movie and the ratings given by a user in order to generate predictions for the movies that the user will like. This prediction is further presented to the user as a recommendation. For example, Netflix uses a recommendation system to predict movies and generate favorable recommendations for users based on their profiles and the profiles of users similar to them. In user-based collaborative filtering algorithm, the movies rated highly by the users similar to a particular user are considered as recommendations for that user. But users' preferences vary with time, which often affects the efficacy of the recommendation, especially in a movie recommendation system. Because of the constant variation of the preferences, there has been research on using time since rating or watching the movie as a significant factor for recommendation. If time is considered as an attribute in the training phase of building a recommendation model, the model might get complex. Most of the research till now does this in the training phase, However, we study the effect of using time as a factor in the post training phase and study it further by applying a genre-based filtering mechanism to the system. Employing this in the post training phase reduces the complexity of the method and also reduces the number of irrelevant recommendations.