Research of Dynamic Text Conversation Extraction
Yating Zhang · Video Engineering · 2011
Automatically analyzing and deriving implied value of knowledge for massive short data has become an urgent demand.Text conversation extraction is proposed for such dynamic data like MSN and QQ to detect intertwined messages that belong to the same conversations,which is a new area of online topic detection and tracing,and playing an important role in the application of information retrieval,text mining,topic detection tracking,etc.In this survey,the necessity and importance of dynamic text conversation extraction are firstly introduced.The main contents and evaluation methods are then analyzed,an integrated research framework is proposed,and the key technologies of framework are analyzed.Finally,the key issues,difficulties,and future directions in the field of text conversation are all summarized.