Text Sentiment Analysis Based On Interactive Attention Network And Application Research

The development of Internet has brought great changes to people’s lives.The Internet is gradually becoming the main carrier of various information in the society.Among them,social network platforms have received widespread attention due to its convenience.Internet users post different comments on various social network platforms.These comment data from Internet users are often mixed with users’ sentiment.Therefore,how to extract core information from these sentiment-colored text data has become one of the research hotspots.Text sentiment analysis can systematically judge the sentiment tendency of these texts with subjective sentiment color,so as to assist users to extract core semantic information.On this basis,this paper conducts a further study.Traditional sentiment analysis methods rely on sequential neural networks.However,traditional sequential neural networks tend to dilute key information in the face of long sequence data.During these years,with the application of attention mechanism and graph neural network in the field of natural language processing,which can improve the ability of the network to process key information.Sentiment analysis methods based on attention networks and graph neural networks have received extensive attention in the field of sentiment analysis.However,existing sentiment analysis methods ignore the role of text syntactic structure,and lack the guidance of professional domain knowledge when dealing with texts in different fields,which restricts the improvement of the performance of the sentiment analysis model.In view of the above problems,this paper proposes a method to integrate external common sense conceptual knowledge and constructing attention network for text sentiment analysis,and the above method is used in the public opinion system,which further verifies the effectiveness of the method.The main work and innovation are as follows:(1)To solve the problem that coarse-grained sentiment analysis cannot fully analyze the text,this paper focuses on fine-grained sentiment analysis,introduces the location information based on syntax,and the performance of fine-grained sentiment analysis model is improved by combining with attention mechanism.(2)To solve the problem that aspect-based sentiment analysis weight dispersion and lack of professional knowledge guidance,in this paper,a semantic graph is constructed based on the knowledge graph,and the graph convolutional neural network is used to enhance the expressive ability of text features.Finally,the robustness and accuracy of fine-grained sentiment analysis model are improved.(3)The method of importing external knowledge and constructing attention network for text sentiment analysis is applied in the actual public opinion analysis system project,The practicability of the method is further proved.