Research on Chinese Micro-Blog Sentiment Analysis Based on Deep Learning

Micro-blog sentiment analysis aims to find user´s attitude and opinion of hot events. Most of studies have used SVM, CRF and other traditional algorithms, which based on manual tagging of a lot of emotional characteristics, but paid a high price. To improve this situation, further studied deep learning and Micro-blog sentiment analysis, and proposed a new technical solution. It firstly crawled some data from Micro-blog through crawler, then after corpus pretreatment, as the input sample of Convolutional Neural Network, and built the classifier based on SVM/RNN, finally judged the emotional orientation of each sentence in a given test set. Verified by examples, experimental results show that this solution can effectively improve the accuracy of emotional orientation, validation result is ideal.