Research On The Construction And Application Of Police-involved Public Opinion Knowledge Map

In the field of public opinion analysis,the traditional public opinion analysis and judgment model is usually constructed based on expert opinions and inference rules.Its disadvantage is that the model construction relies on the knowledge accumulation of industry experts and takes a long time.In today's self-media era,all walks of life pay more and more attention to public opinion,Especially negative public opinion.Using knowledge maps to quickly construct public opinion analysis and judgment models has practical application value.Knowledge maps are divided into general knowledge maps,domain knowledge maps and cross-language knowledge maps.At present,knowledge maps are mostly general knowledge maps,and knowledge maps in police public opinion fields are still rare,It is of great significance to maintain the social security and stability and promote the harmonious development of the society by constructing the police lyric knowledge map and using it in the police analysis system.This paper draws on the existing knowledge map construction method,and proposes the construction method of police sensation knowledge map.The main research contents include:(1)Referring to the entity recognition method in the existing knowledge map,the entity recognition method of police-related public opinion information based on BI-LSTM-CRF is proposed.Aiming at the problem that traditional methods cannot effectively extract the semantic characteristics of police-related public opinion field,the word vector characteristics based on supervised entities are constructed.Experiments show that this method can effectively extract the semantic characteristics of police-involved public opinion field and improve the accuracy and recall rate of entity identification of police-involved public opinion information.(2)Aiming at the problems of wide data sources and complex relationship types of police-involved public opinions,this paper proposes an entity relationship extraction method of police-involved public opinions information based on rules and distance supervision.By analyzing the sentence structure of police-involved public opinions,the rule template of police-involved public opinions information is extracted and used for the entity relationship extraction of police-involved public opinions information.Experiments show that this method improves the recall rate and accuracy of the entity relationship extraction of police-related public opinion information.(3)The cosine distance is used to calculate the similarity between the reference of police-related public opinion information entity and the context of the candidate entity in the known database,so as to link the police-related public opinion information entity to the interactive encyclopedia.In this paper,police-related public opinion information is used to construct the knowledge map of police-related public opinion and apply it to the keyword retrieval system of police-related public opinion information and the ranking of police-related public opinion information keywords.