Research On Intelligent Recommendation Based On Web Data Mining

Internet Web site which contains massive information has a complex structure.Many sites often use the personalized recommendation system to help our users quickly and easily find the information they need to cultivate customer loyalty. In modern information society, personalized recommendation system caters to the great demand for information services and also is an inevitable trend of development. Web data mining technology has its innate advantages when facing massive data and sparse data processing, so Web data mining technology has been more widely studied and applied in personalized recommendation system.This article describes the concepts in personalized recommendation system which based on web data mining technology, and analyzes the relevant technology. For e-business applications, this paper brings up a recommendation system model, discusses in detail the data preprocessing steps, and analyzes the purpose and methods of each step, and gives the algorithm of each step. This article lays emphasis on Collaborative filtering based on user research and project-based collaborative filtering algorithms, and analyzes their advantages and disadvantages of complementary improvements on this basis and implemented projects based on user and collaborative filtering algorithm.On this basis, finally in order to verify the effect of user and project-based collaborative filtering algorithm, we take the well-kno wn database MovieLens in the field of collaborative filtering as the experimental database, and use the project-based recommendation a lgorithm to make analysis and processing. Experimental results sho w that the proposed algorithm based on user and project recommen ded in the algorithm theoretical and experimental results have certai n aspects of performance, is a successful attempt in the personalize d recommendations research field.The future system to be adopted this recommendation and othe r similar recommendation system performance indicators in a compr ehensive comparison, further analysis and assessment of the perform ance of recommendation systems, research to improve the system p erformance of the method.