Research On Academic Social Network Resource Fusion For Subject Demand

With the development of Internet information technology,academic social network as a new practical tool for knowledge exchange and knowledge sharing,has become an important source of information for academic communication and innovation among researchers.However,due to the volume,variety and heterogeneity of academic social network resources,and the lack of semantic information support in most of the current retrieval methods,it is difficult for scientific research users to accurately retrieve valuable information in many academic network platforms.Therefore,it is necessary to reorganize the resources of academic social networks to meet the information needs of scientific users,and to promote the access and effective use of academic social network resources,thereby promoting scientific research innovation and knowledge sharing.As a best practice of Semantic Web,linked data is a structured knowledge organization method,which uses Resource Description Framework RDF and URI to describe information resources in a standardized way.Linked data provides a very effective solution for the semantic description and reorganization of multi-source heterogeneous data under the network environment.Information fusion technology based on semantic association provides an important guarantee for the efficient integration of network information resources and information services in technology,and effectively promotes the development of network information services to intelligent applications.It can well solve the problem of heterogeneity of information resources to construct the related network from the semantic level and reveal the relationship between the content features and the external features of information resources.Thereby,the semantic disambiguation and deep fusion of heterogeneous data can be realized and the recombination and reuse of network knowledge can be promoted effectively.In order to realize the full development and deep utilization of academic social network information resources,and enable scientific research users to carry out academic research efficiently,this paper intends to construct a theoretical system from the perspective of semantic web,and proposes a theoretical framework and solution for the fusion of multi-source,fragmented academic social network resources,using multi-disciplinary methods such as informetrics,knowledge science,and data mining.This paper mainly includes four aspects of research: the quality evaluation of academic social network platform,the distributed data collection from academic resources,the extraction of academic resources' topic features and semantic description of it's external features,and multi-dimensional information fusion based on topics.Firstly,the quality of information resources of eight alternative academic social network objects was evaluated using the multi-objective evaluation method of evidential reasoning,and the evaluation result was used to select experimental data.The selected data is collected by distributed data collection to obtain experimental data sources.Secondly,in the dimension of content features,the LDA topic model is used to cluster the topics of heterogeneous academic resources,and Silk Workbench,a semantic association discovery platform,is used to reveal its potential topic associations.In the dimension of external features,the RDF semantic description framework was used to transform the format of the external feature information.Finally,Silk Workbench is used to automatically discover the internal and external features of topic-oriented academic social network resources,so as to realize the deep fusion of multi-source academic social network information resources.