Research On The Personalized Recommendation Algorithm Of Social Network

With the popularity and application of social networking,a large number of users are flooding into social networks,and then it causes the problem of "information overload" in the social network.In order to help users search for information that they may be interested in,and improve the user experience,the personalized recommendation algorithm was proposed.In order to solve the problem of deviation between the recommended quality and the actual situation and the quality of recommendation is reducing which was caused by the lack of community number,the main research work of this paper is as follows:1.The traditional personalized recommendation algorithm cannot make recommendations for users dynamically and using display interactive behavior to collect information is fussy,and then a good friend recommendation algorithm based on interest transfer is proposed.In order to be more realistic,the algorithm introduces the Ebbinghaus memory curve to simulate the oblivion of the human brain and adds the theory of interest transfer.In view of the non-standard recommendation problem of the user's hesitation time,the user similarity calculation method is redefined by the user browsing time.Finally,the experiment is carried out through the real data in the social web site to verify the effectiveness of the algorithm.2.In view of the lack of community number in community friends recommendation algorithm,which leads to the reduction of recommended quality,a good friend list expansion algorithm based on similar community is proposed.This method first selects the authoritative users from the community,and then uses the historical micro-blog text of the authoritative users to dig out the key words of the community,then the similarity and the relevance of the keywords is calculated to find the similarity community.Finally,the similarity community is fused to make recommendation for users.Experiments on real data in social networking sites show that the algorithm is more suitable for small social networks with a small number of people,and the accuracy and recall of the algorithm are improved compared with the two methods of extended friends list mentioned in the CUPC algorithm.The experiment results show that the two algorithms proposed in this paper are better than the original algorithm in terms of user similarity,accuracy and recall.The use of this algorithm in the social network is of positive significance for the user to improve their user experience.It has a positive significance for the operators to improve the user stickiness and enhance the income,and it also has a certain theoretical significance for the research of personalized recommendation algorithms in other fields.