Sentiment classification is the main and popular task in the field of sentiment analysis. Most of the existing researches focus on how to extract the effective features, such as lexical features and syntactic features, while limited work has been done on the extraction of semantic features, which can make more contributions to sentiment classification. This paper presents a method for sentiment classification based on word2vec. Word2vec is a tool, which establishes the neural network models to learn the vector representations of words in the high dimensional vector space. So it can extract the deep semantic relationships between words. In this paper, firstly, we cluster the similar features together using word2vec. And then we use word2vec again to learn the word representations as candidate feature vectors. After feature selection, the SVM perf package is adopted to train and classify the comment texts. To conduct the experiments, we collect a large number of Chinese comments on clothing products as data set. The experimental results show that the accuracy of sentiment classification is over 90 percent, which proves the effectiveness of proposed method for Chinese sentiment classification.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.