Sentiment Classification of E-Commerce Product Quality Reviews by FL-SVM Approaches

摘要

^{_In order to improve the accuracy of sentiment classification for e-commerce products quality reviews, this paper proposed the FL-SVM approaches which use the fuzzy theory measure the level of emotional polarity of sentiment words to construct the fuzzy sentiment dictionary based on the How Net, the support vector machine (SVM) is adopted to construct the classifier model for evaluating the sentiment tendency of e-commerce product quality reviews. Compared to the NB, kNN and SVM algorithm, the experiments on different data sets show that the FL SVM approaches can achieve the better sentiment classification accuracy increased by 1 % ~ 3%, which verifies the effectiveness and robustness of the proposed approaches.}