Information Dissemination Model Based On User Behavior In Microblog

Microblog as a major platform of information sharing and interaction in socialnetwork, the dissemination of information in time and space has changed obviously as itemerged. User is not only recipient of the information, but also disseminator of theinformation on microblog platform, whereby, there are two ways of disseminatinginformation, one is tweeting an original tweet, and another is retweeting other’s tweets.The dissemination of microblog information relys on users’ retweeting behavior. Throughresearching dissemination law of microblog information and the intrinsic relationshipsamong users’ retweeting behavior, users’ relationship and microblog contents, we foundthe core impact factor in the process of the dissemination of information, thus it couldpredict and control the scope of information.Based on Sina Weibo, we acquired information of users, relationships of users andmicroblogs by calling open platform API and crawling microblog webpages, thenextracted and analysed features of users’ and microblog contents’ information.Formalization definitions of the retweetability, the influence of key node and the basicconcept of information dissemination capacity were provided, and metric method of theirswere depicted. At last we proposed one way of building the relationship network ofmicroblog information dissemination.User receives a tweet, performs retweeting decision after understanding its content.The decision depends on user’s personal preference and understanding of the information.Thus retweetability can be decided after analyzing and computing. To predict whether userretweets a received tweet, further determine the dissemination scope of microblog, webuilt a microblog information dissemination model based on user behavior analysismaking use of the retweetability P of user’s tweet and impact factor of key nodes. Then doa detailed description of model core idea, construct the corresponding algorithm. Finally,the algorithm was implemented through Java.In order to verify the scientificity and efficiency of model, we conducted simulation and verification experiment of model. The result shows that the model can predict atweet’s retweeted number.