Computer Science · Information Systems
ACIRS
Mahamudul Hasan, Farzana Aktar Lubna, Sadia Chowdhury, Nusrat Jarin Haque, Towsif Ahmed Omi
See on the mapAbstract
Recommender system is that which collect users preferences and suggest the user some similar item that the user might like. It increases our interest in e-commerce sites by suggesting items. Recommender system suggests the appropriate items to the users based on his previous inclination. In clustering techniques, similar users or products are clustered together that can be used for a product recommendation. But it is difficult for an individual clustering algorithm to find a correlation that can help in recommendation. An item-based recommendation system has been proposed here based on clustering algorithms that can find the similarity between items. Item-Item based clustering algorithm first analyzes the user-item rating matrix to find out the relationship between different items and then use these relationships to compute recommendation for users. Finally, our experimental result provides good performance in terms of mean absolute error, precision, recall, and f-measures.