A new similarity coefficient for a collaborative filtering algorithm

Recommender systems give the opportunity to present automatically personalized content across many digital marketing channels to visitors depending on visitor movements on the site. In recent years, there has been a lot of interest in e-commerce companies in order to offer personalized content. So, recommender systems become very popular and many studies have been done in this regard. New works are being done day by day to improve the results. In this paper, we propose a new memory-based collaborative filtering algorithm. Calculation of similarities between items or users is a critical step in memory-based CF algorithms. Therefore, we proposed a new function for calculation of similarities based on user ratings. In this study the more similar the user's pleasures are, the more similar it is to the products the users choose, is adopted. The adopted idea in this study is that the more similar the user's pleasures are, the more similar products are chosen. We estimate the degree which a user is interested in X product. To do this, we find other users who are interested in product X and calculate the similarity ratios of those users to the user. We tested our algorithm in MovieLens 100K dataset and compared to other similarity functions. We used MAE and RMSE measures in our experiments.