Research On The Construction Of Emotional Relevance Data In Chinese Online Reviews

The World Wide Web has changed the way of communication of people,especially the development of social media,that makes people publish and di sseminate information freely.In this information,in addition to the factual obj ective information,opinion and emotion and other subjective information is r apid growth.More and more users share their evaluation and feelings of event s on the network.Increasing number of users and emotional information has attracted the attention of scholars and related industries.However,now the fie ld is generally faced with the problem that the insufficient resources of emoti onal analysis,at the same time,the resources are independent of each other,p oor sharing.Worse,the heterogeneity of the emotion expression and model m akes it impossible to connect and reuse different resources.Therefore,this pap er uses the linked data of the semantic web technology to describe the emotio nal semantics in the text,through constructing the emotion metadata ontology,releasing affective association data,realize the sharing of resources and emo tional semantic query.The research contents of this paper are as follows:The first chapter mainly introduces the significance and research backgr ound of emotion correlation data,and analyzes the current situation of researc h in this field.The second chapter mainly introduces the technical overview of the sem antic network,the resource description framework RDF,the ontology and ont ology description language OWL and the related concepts of the linked data,and makes technical preparation for the construction of emotional association data.The third chapter describes the emotional metadata through the OWL on tology description language.This study mainly uses the widely used metadata ontology SKOS,PROV,FrameNet Ontology and an WordNet Affect,mutual mapping,to achieve the relationship between different ontologies.The fourth chapter is based on the ontology constructed in Chapter 3,collects and annotates the emotional semantic resource data,converts the data i n RDF format,and completes the association between the data sets.To achiev e the release of emotional association data.In the fifth chapter,published,browsed and inquired association data on the basis of the structure of association data.Jena Fuseki has completed the c onstruction and querying of the online commentary emotion correlation data,and realized the interaction between the resources.Chapter 6 summarizes the research work and innovation of the paper an d puts forward the direction of further efforts.The results of this paper provide a machine-readable and understandable semantic resource for emotional analysis and semantic mining,which provid es an effective knowledge structure and data foundation for emotional semant ic reasoning and intelligent application.