Temporal Analysis of User Behavior and Topic Evolution on Twitter

摘要: We investigate the temporal aspects of user behavior in this paper and relate it to evolution particular topics being discussed on Twitter OSN. Studies have shown that a small number frequent users are responsible for maximum percentage tweets. further hypothesize deviate from their usual tweeting hours when major event occurs. With these as our underlying concepts, we introduce new metric called "tweet strength" gives more weight tweets by who general tweet less or those do not usually at given time. study set through lens strength try identify classes driving popularity different times. also word-of-mouth diffusion mechanism network defining "copying" behavior. When follower same topic user, is said copied. make definition time-dependent imposing thresholds it.