Study on the Subjective and Objective Text Classification and Pretreatment of Chinese Network Text

Subjective and objective text classification is widely used in product reviews, video reviews, social public opinion analysis and micro-blogging attitude analysis. To solve the existing problem of network text formalization in subjective and objective text classification, a machine learning classification method based on network informal language (NIL) is proposed. Firstly, a network informal dictionary is constructed by writing a web crawler to collect informal words which can be divided into two categories: typical type and fuzzy type. Then, different methods are put forward to formalize the informal network text based on the two types of informal words. Finally, we adopt the Native Bayes classifier and Sequential Minimal Optimization classifier to distinguish subjectivity and objectivity of the text. The experimental results reveal that the method we proposed can improve the accuracy of subjective and objective text classification.