Research on the Construction Framework of Multi-Modal Knowledge Graph for Social Network Public Opinion
[Research purpose]The development of information technology has enriched the communication methods of social media users.The construction of a multimodal knowledge graph for social network public opinion has important practical significance for public opinion governance on the internet.[Research method]Based on the semantic complementarity of multimodal data,various heterogeneous data fusion methods,such as entity attribute association,image(video)textual description,image(video)attributes,and image(video)as-sociations,are discussed.Based on these,a framework for constructing a multimodal knowledge graph for social network public opinion is proposed,and the problems and challenges in multimodal semantic understanding,multimodal entity alignment,and multimodal knowl-edge representation are analyzed.[Research conclusion]A framework for constructing a multimodal knowledge graph for social network public opinion based on multimodal knowledge fusion is proposed,providing valuable reference for the increasingly diverse interaction methods in public opinion governance on social networks.
social mediamultimodalmultimodal knowledge graphmultimodal dataonline public opinionpublic opinion govern-ancesentiment analysis
国家科技期刊平台
NSTL
万方数据
维普
