Analysis of Twitter Trending Topics via LinkAnomaly Detection

Abstract

Most of us be aware ofrising themes signaled by way of social disorderswith these sites. In exact, we pay attention tothe mentions of users-links in between end userswhich can be generated dynamically (intentionallyas good as unintentionally) by way of responds,retweets, additionally to mentions. The probabilitymodel of the mentioning behavior of a social enduser captures both the number of mentions per postas well as the frequency of mentionee. We thenaggregated the paradox rankings from numerous endusers and we show that we may realize thetrending themes easily utilising the reply/point outrelationships in twitter posts. In this paper, themixture of point out-anomaly model with termfrequency ways is proposed. We illustrate ourprocess on the datasets received through twitter