Abstract— Understandnatural language texts is the need for machines highlights explosioninformation short text refer to text with limited context. Many applicationsjust like web search and micro blogging services etc. having need to handlelarge amount of short text data. Understanding short text data bring tremendousvalue in social media mostly. Existing system having the challenges short textsdo not always observe the syntax of written language texts. As a resulttraditional natural language processing tools, ranging from part- of-speechtagging to dependency parsing can not be easily applied. Second statisticalsignals are not sufficient for text mining. Short text is more ambiguous andnoisy, and are generated in an enormous volume. Hence short text handlingbecomes complicated. I think semantic knowledge is required to understand shorttext in better way. In this paper system use semantic knowledge which isprovided by a well known knowledgebase and automatically harvested from a webcorpus. Proposed system overcomes the challenges by using text segmentation, partof speech tagging and concept labeling. our knowledge intensive approachestowards of both hot event evolution anddiscovering semantics of short texteffectively and efficiently.
Keywords: Short text understanding, text segmentation, type detection,concept labelling, semantic knowledge, Micro blogging, Event Evolution, User Topic etc.