The Design And Implementation Of Sentiment Analysis Service Based On Lstm And Attention Mechanism

Sentiment analysis or opinion mining is the research of people's sentiment tendency to a certain entity.In recent years,with the rapid development of the Internet,sentiment mining of social networking texts has also become a hot topic.However,most of the sentiment analysis methods used in the whole Chinese environment are mostly out of date methods,such as sentiment lexicon based method and SVM based method,and they can not get good performance when they need a lot of manual work.Most of them only stay in the research of the model,and have not built it into a practical end-to-end sentiment analysis service system.This thesis first summarizes the related technologies of sentiment analysis services,including the deep learning related technologies,such as words vector,recurrent neural network,the long short term memory mechanism,and the attention mechanism,and the end to end service system construction related technologies,such as Docker and Kubernetes technologies.On the basis of studying the application of deep learning in the field of Chinese sentiment analysis,this thesis constructs a deep model of sentiment analysis based on Bi-LSTMs and Self-Attention mechanism,and then saves and deploys the trained model in the Kubernetes cluster environment to build the end to end sentiment analysis service system.The deployment service consists of two parts:the server and the client.The client accesses the server's program and model through the port exposed by the server,and the user accesses the service through the exposed port of the client.The end to end sentiment analysis service system is widely used in reality.On one hand,it can be applied to scientific research in related fields,such as semantic analysis of text.On the other hand,it can also be applied to business fields.It can be used to analyze the opinion tendency of a certain entity in the network and then assist business decision.