Improve recommender systems using certain formulas and fuzzy concept networks
Elaheh Kafshi Tagiabadi, Mehrdad Jalali · 2013
Personalization systems are presented to users to improve suggestion when users are surfing on the net. In order to create accurate model from user webpage contents can be used. Recently some researches were done for adding semantic to user behaviors. However, all users don't show all their interest, but the system must scrutinize user behavior and suggest best prediction for his interest. The best result of this paper is presents a mechanism to improve both user behaviors in web site and exploits fuzzy concept for building automatic model from user interest automatically. Also when the system proposed an item to users in order to calculate item similarity, use collaborative filtering in which to predict user interests uses enriched prediction formula. The results show that the proposed method provides better prediction than traditional methods.