Abstract:
The tremendous growth in the amount of available information and the number of visitors to Web sites in recent years poses some key challenges for recommender systems. Recommender systems form a specific type of information filtering (IF) technique that attempts to present information items ( movies, music, books, news, images, web pages, etc., ) that are likely of interest to the user. They produce high quality recommendation and perform many recommendations per second for millions of users and items and achieving high coverage in the face of data sparsity. Most systems are implemented using the k-nearest neighbor collaborative filtering but have some weakness in searching on the Web. To address these issues, item-based collaborative filtering techniques have been explored. Firstly, item-based techniques analyze the user-item matrix to identify relationship between different items and then use these relationships to compute indirectly the user’s profile to some reference characteristics, and seek to predict the ‘rating’ that a user would give to an item they had not yet considered.
