Sentiment analysis method for e-commerce review on weak-label data and deep learning model

For inaccurate weak-label data of e-commerce reviews, the traditional manual labelling method is time-consuming, and it is necessary to solve the problem of polysemy and imbalance in Chinese reviews to improve the performance of sentiment analysis model. This paper collects agricultural product reviews on Jingdong platform. Firstly, the improved SO-PMI method is used to construct a domain sentiment dictionary, by combining the review sentiment tendency calculated by the dictionary with the weak-label data of user ratings, an unsupervised generation of high-quality training sets is realised. Secondly, two basic learners, Bidirectional Long Short Term Memory (BiLSTM) and Convolutional Neural Network (CNN), are combined in the sentiment analysis model, and the character, word, part-of-speech vector features are extracted in parallel. In addition, an attention mechanism is embedded in the channel, and using Focal Loss during model training process. The experimental results show that the accuracy of the method proposed in this paper reaches 97.34%, which is 4.64% higher than that of directly using weak-label data for training. Compared with single-channel CNN and BiLSTM model, the accuracy is improved by 1.55% and 0.99% respectively. Therefore, this method improves the accuracy of sentiment analysis of e-commerce reviews.