Evolutionary Modeling And Simulation Reaserch Of Personalized Recommendation Based On The Social Network Model

The progress of Internet and social media promote the development of social networks, and change the mode of information communication and dissemination.in which user and information recommendation as an important part of the social network, but also improve the user-stick degree of social platforms. More and more e-commerce, social networking platform based on user interests dig, so to recommend products or information, to improve their profitability. However, people are interested in how to find data from a large number of network users and the information is very difficult, and our search engine to all users are the same results, but can not meet the individual needs. More and more recommendation system appeared in front of us, there are a variety of defects, such as cold start, sparsity and other issues, and no combination of social networks that can not be accurately personalized recommendations.This paper studies how provide micro-blog users with efficient and accurate personalized recommendations, and the main purpose of micro-blog users is to find the information and users they are interested in. Microblogging is different from Renren, Kaixin network and other social networks, it is a directed network, because of mutual concern between the user and forwarding and information dissemination. In this paper, in order to dig more microblogging personalized recommendations.Firstly, microblogging information dissemination mechanism, mainly using the improved SIR Epidemic Model model to build microblogging information dissemination model, access to information dissemination process, again microblogging users are divided into three categories: the node is not in contact information microblogging, microblogging information dissemination and receiving nodes but not to disseminate tweets nodes and use microblogging API to fetch related data simulation. Experimentally known, not only through the microblogging forwarding, comments, thumbs and other ways of information dissemination, but also through external media, such as the system recommended by a friend. Secondly, attention from social networks- to be concerned about microscopic model to study the mechanisms of attention microblogging network, the main factors driving attention to the relationship between the formation of micro-blog users. Again, collaborative recommendation algorithm combines microblogging attention- to be concerned mechanism and information dissemination mechanism microblogging personalized recommendation to study, and according to some of the existing problems, the improved matrix factorization model binding rules passed, similarity propagation who, like the three recipients of the model rules to guide learning. Finally, the use of data to validate the model, in order to verify the validity of microblogging personalized recommendations Mechanism.Microblogging personalized recommendation applies not only to the recommended friends and microblogging information in microblogging social networking, it can also be applied to e-commerce. Combining social networks attention-concerned and information dissemination mechanism for e-commerce users recommend the appropriate goods and microblogging friends, via the corresponding experimental verification, compared with the traditional recommendation of personalization, the proposed personalized Weibo Suggest mechanisms to improve the effect of the recommendation.