Virtual Group Movie Recommendation System Using Social Network Information

Abstract

AbstractRecommendation systems (RS) are software tools and methods designed to give recommendations to support customers in different decisions in terms of what items to buy, music to listen to, news to read, and so forth. Most recommender systems recommend items in terms of individual user likings and group recommender systems recommend items taking into consideration the likings and personalities of group members. To generate effective recommendations for a group, the system must satisfy, to the greatest extent possible, the individual interests of the group members. With the social networks, it is possible to recommend to a virtual group thus this study endeavors to develop a virtual group recommender system prototype using a model-based matrix factorization algorithm of collaborative filtering technique then popularity vote for virtual group. A publicly available dataset was used in this study. The results of the prototype showed the proposed collaborative filtering algorithm for prediction of user rating preferences demonstrated a good mean average error (MAE) of 0.70 and root mean square error (RMSE) of 0.89. Virtual groups of social networks user were then formed using the popularity vote algorithm and the results were plausible. This type of recommendation to a virtual group also enables members of the group to have something to talk about on the social network.KeywordsMean absolute error (MAE)Root mean squared error (RMSE)Recommendation system (RS)Collaborative-based filtering (CF)Content-based filtering (CBF)