Graph-Based Chinese Text Categorization

Most text categorization methods are based on the vector-space model which ignores the structural information, such as the order and co- occurrence of words within the text. In this paper we describe a novel approach of text classification using graph matching. To begin with, we use weight method to select the relevant features to construct the graph. Then we present an improved graph-based text representation model and describe a learning algorithm for building category graphs from the training set. Finally we use the graph matching algorithm we proposed to predict the category of the texts in the testing set. The result shows that the graph matching approach can outperform traditional vector-based NB methods in terms of both accuracy and spend time.