Chinese Weibo Sentiment Analysis Based On Multiple Sentiment Dictionaries And Deep Learning

In recent years,Chinese media platforms represented by Weibo are constantly being integrated into people's lives.People express their opinions,feelings and other subjective information on these platforms every day.How to extract valuable emotional information from these information and use it Using it is called sentiment analysis.This article takes Chinese Weibo as the research object,and conducts sentiment analysis research on Chinese Weibo from the method based on sentiment dictionary and the method based on deep learning.The research results prove that both methods have a good effect on sentiment analysis of Chinese Weibo The main research contents are as follows:(1)First of all,according to the characteristics of existing sentiment dictionaries and Weibo,it can be found that the existing sentiment dictionaries can not meet the sentiment analysis of Weibo,so this article specially developed and expanded six sentiment dictionaries,including original sentiment dictionaries,negative words and double Negative dictionary,adverb dictionary,conjunction dictionary,emoticon dictionary and Chinese Weibo new word sentiment dictionary,among which the construction of Chinese Weibo new word sentiment dictionary is a key point of this method,which is constructed based on the improved PMI algorithm.Secondly,assign a sentiment value to each word in each dictionary,and then further analyze the semantic rule set between Chinese microblog texts,and introduce inter-sentence analysis rules and sentence pattern analysis rules into the sentiment analysis of Chinese microblogs.Improve the accuracy of sentiment analysis of Chinese microblog.Finally,based on the method of multiple sentiment dictionaries and semantic rule sets,this paper proposes a Chinese Weibo sentiment calculation algorithm that combines complex sentences to single sentences,then single sentences to words,and emojis.According to this algorithm,the Weibo data set is divided into three categories:positive,negative and neutral,and then through three sets of comparative experiments:a method based on the original sentiment dictionary,a method based on multiple sentiment dictionaries,a method based on multiple sentiment dictionaries and a rule set The sentiment analysis of Weibo's method shows that the proposed method based on multiple sentiment dictionaries and rule sets is superior to the first two methods in terms of accuracy,recall and F value.(2)In order to study the effectiveness of sentiment analysis on microblogs using deep learning methods,this paper proposes a combination of the two based on the advantages and disadvantages of the methods based on CNN and LSTM based on long and short-term memory network The model is the CNN-LSTM model.This model first uses the convolutional neural network CNN to extract features from the microblog,and then uses the long and short-term memory network LSTM to predict emotions on the microblog.Through experimental comparison,it is found that the sentiment analysis of this model on Weibo is greatly improved in accuracy,recall and F value compared to the single CNN model and LSTM model.Figure 31 table 6 reference 80...