A COMPARATIVE STUDY ON RECOMMENDATION SYSTEM USING HYBRID APPROACH

Recommendation is a wide used technique to guide user to settle on the proper product on-line because it becomes the essential feature for a better E-commerce. Many recommender applications use item to item collaborative approach, i.e., overlapping previous user's data with new user's data and recommending remaining items. An item can be recommended to the user based on user's interests, favorite subjects, occupation of the customer, person's gender, age, etc. This literature survey tries to explain about item to item collaborative approach along with new features which improves the performance of the system. Not only in E-commerce, in many web applications recommendations play an important role such as book recommendation for a library, research paper recommendation based on previous searches while searching for a paper, dynamic recommendation system using web data mining and on social websites.