期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2017
卷号：6
期号：2
页码：1996
DOI：10.15680/IJIRSET.2017.0602101
出版社：S&S Publications
摘要：Twitter has turned out to be one of the biggest stages for clients around the globe to impart anythingoccurring around them to companions and past. A bursty point in Twitter is one that triggers a surge of pertinent tweetsinside a brief timeframe, which regularly reflects essential occasions of mass intrigue. The most effective method to useTwitter for early identification of bursty points has along these lines turn into a critical research issue with enormouscommonsense esteem. In spite of the abundance of research work on point demonstrating and examination in Twitter, itremains a tremendous test to recognize bursty subjects continuously. As existing strategies can barely scale to deal withthe errand with the tweet stream continuously, we propose in this paper TopicSketch, a novel outline based subjectmodel together with an arrangement of methods to accomplish constant location. We assess our answer on a tweetstream with more than 30 million tweets. Our investigation comes about show both proficiency and adequacy of ourapproach. Particularly it is additionally exhibited that TopicSketch can conceivably deal with many millions tweetsevery day which is near the aggregate number of day by day tweets in Twitter and present bursty occasion in bettergranularity.
关键词：Tweets; Data Stream; Bursty Topic Modelling; Practical Exposure; Twitter Analysis.
