A Text Network Representation Model

Text representation is the basis of text processing. Most current text representation models ignore the words´ inter-relations, which result in the loss of textpsilas structure information. This paper proposed a novel text representation model, which uses lexical network to represent the text and retains the text´s structure. According to the different levels of words´ inter-relations, co-occurrence network, syntactic network and semantic network are introduced. To evaluate the representation ability of text network representation model, we investigated the applications of text network to two language processing tasks including unsupervised keyword extraction and text classification. The experimental results show how to use it for natural language processing successfully.