Abstract: As a new field of sentiment analysis, text mining can be used to classify and summarize online users’ product reviews. The summaries and classification help provider to improve service and product quality, and also provide buyer advicse for other consumers. The paper proposes a sentiment analysis method based on sentiment words extraction and LDA feature representation, for online products’ reviews making binary classification. The processing steps are as follows: extract sentiment words from the preprocessed text using the manually created sentiment dictionary; create the topic subject distribution of documents using the LDA model; take comment-subject distribution as feature; do classification based on the SVM classifier. Experiments show that, the proposed method has excellent effects of review of judgments classification.


Key words: sentiment analysis, sentiment words extraction, latent Dirichlet allocation, topic model, support vector machine
