Research on Public Opinion Analysis Methods Based on Knowledge Mapping and Deep Learning

In order to effectively respond to public opinion events, it is necessary to integrate various information sources and construct a comprehensive and accurate knowledge base. Knowledge graph, as an emerging data structure, can be a good solution to this problem. This paper proposes a method that combines the advantages of knowledge graph and deep learning technology to improve the accuracy and professionalism of public opinion analysis. In the study, police-related public opinion data from different datasets are fused and de-duplicated through data preprocessing, and core words are extracted for theme mining and sentiment analysis. The optimisation algorithm model is incorporated into knowledge graph entities, and machine translation and entity recognition techniques are used to improve the reliability and pertinence of the algorithm. Experiments show that using the knowledge graph as the input and constraint of deep learning can improve the efficiency and accuracy of the algorithm. The research results have certain guiding significance for improving the professionalism and accuracy of police-related public opinion analysis.