Cross-language sentiment classification based on Support Vector Machine

This paper presents a cross-language sentiment classification method that based on the Support Vector Machine model. The method is based on the research about English training corpus, first of all, we use statistical methods extract feature words in English, and use machine translation tools build an “English-Chinese” feature word bank. Then, we put forward a feature word weighting method which combined TF-IDF with sentimental intensity of sentiment words, after that, we constructed a vector space model. Finally, we optimized the Support Vector Machine classification model by using a joint training set. Experimental results show the effectiveness of this method.