11-11-2014, 10:23 AM
Abstracts: Due to the huge amount of information available online, the need of personalization and filtering systems is growing permanently. Recommendation systems constitute a specific type of information filtering technique that attempt to present items according to the interest expressed by a user. Commonly online recommenders are employed for e-commerce applications or customer adapted websites. There exist two basic types of recommendation techniques, namely content-based filtering and collaborative filtering. Content-based filtering methods examine items previously favoured by the actual user and collaborative filtering computes recommendations based on the information about similar items or users. We have explicitly collected the data of the users Like-Dislike of Perfumes, their Perfume Features. We have implemented k-nearest neighbourhood to find nearest k users of particular user and matrix factorization of collaborative filtering and also content based to find item similarity for new user problem. So we are using hybrid approach for recommendation. We have implemented all the matrix operations in MATLAB.