Text Classification Based On Hidden Markov Model And Semantic Fusion

Text classification is the key technology of organizing and managing large-scale text data,and it is also an important task in natural language processing.There are many mature automated classification methods in statistical model,but these methods lack information related to text semantics and syntactic structure.This paper proposes a text categorization method which based on hidden markov model and semantic fusion.For the classical text classification method,the feature dimension is too high and the semantic phenomenon of the characteristic words is neglected,and the text classification model of the hidden markov model is proposed by using the information gain which extracts the semantic semantics of the feature word through word2 vec model.In order to solve the problem of text syntactic structure loss,a text classification method which is based on hidden markov model and topic model is proposed.Hidden markov model extracts syntactic structure,the topic model extracts feature semantic information,The results of the hidden markov model is great comparable to the classical model by the experimental evaluation.