Research On User Trust Model Of Microblogging Social Network

With the rapid development of social networking services, people more and more dependent on social networks for information exchange, social network users increases rapidly, also accompanied by a large number of malicious users and normal users find other trusted users becomes more and more difficult, the trust problem between users is becoming more and more serious, which makes the problem of trust research become more meaningful, the existing research the user trust application for recommendation system, some research pointed out that it is possible to extract user information and user trust information from the social network, and then in the recommendation system using these information and will make recommendations more effective.The main work of this paper: the research between adjacent users of the microblogging social network trust measurement, through the research and analysis of user behavior and social network, proposes a neighbor users micro blogging social network trust measurement model,called I-Trust model, the model is through the analysis of source of trust about microblogging social network users, the trust is divided into three dimensions to measure: the initial information consist of similar trust, which is similar to user’s interest; information interaction between users, consist of two users interaction trust, it investigates the frequency of interaction between users; the reputation of the users in the microblogging social network consist of the target users global trust, the user’s fans number, the number of microblogging be commented or retweeted. Among them, the interactive trust reflects the asymmetry of trust, and it is more realistic to reflect the characteristics of trust in social networks and real life. For the three dimensions of trust degree calculation, first of all, the article gives the definition, respectively of the these three dimensions of analysis of the existing algorithms, in existing algorithms selected for micro blogging social network algorithm, improved an existing interaction trust degree algorithm, from the formed a new based on micro blogging social network adjacent users trust value measurement model I-Trust.Using the I-Trust model in the millions of users of Tencent microblogging data set on the experiment conducted to verify the validity of the model, the analysis of the situation of trust in the microblogging social network. Finally, the trust value of the three dimensions calculated by the I-Trust model is calculated, and the user recommended experiments are carried out to investigate the performance of the trust value of each dimension in the user’s recommendation problem.