Personalized Recommendation System Recommendation Engine Prototype System

Personalized recommender system is one of the most popular technologies in E-business area. Its main function is providing production information and suggestion by using E-business website. It can help customers to decide which to buy and help them finish the shopping program as a role of sellers. Personalized recommender system can not only help the customers, but also promote the customers'satisfaction degree to the business activities.In today's nation, the practice standard of E-business technology in this area is not high enough, lack of personalization, method of recommending being simple, not automatically. In spite of the importance of recommender system, it still contains lots of weakness. One of the factors is the matter of sparseness, extents, real-time and exactness which arise with the rapid increase of information.This research focuses on exactness and intelligence problem in recommending system, improves the recommending engine by regulating and filtering rules in recommending system, as a result it solves the exactness and intelligence problem. This study's key point is doing research on CAP and FP-Tree arithmetic, improving it as well as implying it. This paper also put improved system into practical use in assurance area, got valuable results.This paper's research is based on relation arithmetic recommending system, studied the system and Apriori arithmetic as well as the recommending engine. Paper also studied structure of recommending engine, found rule base as an improved point, testified the feasibility of using CAP and FP-Tree arithmetic, designed application environment. This research is using a new method to solve a popular problem.This study could help E-commerce recommending system applied in other industry area besides retail section. It could improve the efficiency, profit and core competence of an enterprise.