Research on Improving the Real time of E-commerce Personalized Recommendation System

摘要

In order to solve the problem that E-commerce personalized recommendation system cannot act in real-time on user's access, page increasing or decreasing and site structure changing, this paper proposed a new model based on page clustering. The model consist with the offline and online parts, and used a directed weighted graph to represent the Web access model, then defined a new degree of user access interest, especially proposed three algorithm: the clustering algorithm, the updating clustering sets algorithm and the recommending algorithm. Finally, this paper verified the feasibility of the model by experiments with Visual Studio 2005 software. Experiment shows that the new model can solve the real-time problem of existing E-commerce personalized recommendation system effectively.