Research On Key Technologies Of Personalized Recommendation Based On Clustering

At present,all kinds of information has become a part of our daily life.How to quickly find the information we want is an urgent problem needing to solve.Portal provides people with information classification.Search engines provide information to the user by interacting with the user.However,due to the rapid growth of information,the portal site has been unable to put all the information layout to the site's home page classification.If the user does not know what input information,search engines will be powerless.So both are not a good solution to this problem.Personalized recommendation technology can recommend information to users in the active,do not require the user to take the initiative to participate,and the recommended process is transparent to the user.At the same time,the personalized recommendation can respectively with an organic combination of portals and search engines,further play to the potential of portals and search engines.This paper first introduces some data mining techniques which are closely related with the recommendation technology,expounds how to use the data mining technology in the recommendation system,and then introduces the most widely used recommendation technology.This paper implements the combination of data mining algorithm and recommendation algorithm.Solving the problem of high dimension of vector in the process of user clustering by dimension reduction technique.At the same time,the data filling technique solves the problem that the user is sparse in collaborative filtering recommendation.In the experiment we tested different data,which includes different number of users and items.The comparison of the experimental results shows that it can effectively improve the accuracy of the recommendation,and can adapt to large scale data processing.