A Review Paper on Improvised Method for Tweet Segmentation Using Named Entity Recognition

Twitter has become one of the most important channels of communication with its ability to provide the latest and latest information. Given the extensive use of Twitter as a source of information, touching an interesting tweet for users among a bunch of tweets is a challenge. In this paper, we propose a new framework for batch tweet segmentation, called EnhancedSeg, by dividing the tweets into meaningful segments. Semantic or contextual information is well preserved and easily extracted by downstream applications. Enhance Segmentation finds the optimal segmentation of a tweet by maximizing the sum of the membership scores of its candidate segments. The sticky score considers the probability that a segment is an English expression (ie, a global context) and the probability that a segment is an expression in the tweets batch (that is, The local context). For the latter, we propose and evaluate two models to derive the local context by considering the linguistic characteristics and the temporal dependence in a batch of tweets, respectively.