Personality-Based Matrix Factorization for Personalization in Recommender Systems

Recommender systems are one of the extensively used knowledge discovery applications in database techniques and they have gained a lot of attention in recent years. These systems have been applied in many internet-based communities and businesses to make personalized recommendations and eventually in order to obtain higher profits. The core entity in recommender systems is ratings from users to items. However, there are many auxiliary pieces of information that can be used to get better performance. The personality of users is one of the most useful information that helps the system to produce more accurate and suitable recommendations. It has been proved that the characteristic of a person can directly affect his or her behavior. Therefore, in this paper the personality of users is extracted and a mathematical and algorithmic approach to utilize this information is proposed. The base model that is used is matrix factorization, which is one of the most powerful methods in recommender systems. Experimental results on MovieLens dataset demonstrate the positive impact of personality information on the matrix factorization technique and also reveals better performance by comparing with the state-of-the-art algorithms