A Study On Chinese Text Classification Based On Semantic Graph

With the rapid development and popularity of Internet technology,people need todeal with the daily information increasingly.Facing a flood of information resources, itis difficult to find the real needed information resources quickly and effectively. Thetext classification,as a basic technology of information filtering, information retrieval,search engines, digital libraries and other areas, has broad applications.In this study, it applies the graph structure theory to the model representation,tosolve the problem of missing semantic information in the traditional statistics-based textrepresentation methods.It proposes a new Chinese text classification which based onsemantic graph structure.Firstly,it analyzes the vector space model,and then proposesthe text representation of semantic graph structure for the shortcomings of expressingsemantic information effectively.It solves the problem of information drain in theprocess of text representation,through the semantic graph structure.Secondly,itintroduces the method of text classification based on semantic graph structure.Finally,the text classification system--RCSGC,based on semantic graph structure,was defined.The result of experiments shows that, compared with the traditional vector-basedSVM (Support Vector Machine) and other statistics methods, RCSGC method wasbeneficial to the text representation of the semantic information.And RCSGC methodhas better performance than the traditional SVM method.