Research On Equivalent Entity Network Based On SameAs Interlinking

The Web of Data is growing rapidly,with an ever-expanding set of inter-connected datasets depicted in the Linking Open Data(LOD)cloud diagram.The core of Linked Data is "Linking." The Linked Data can enable users to travel between different datasets,helping users to expand the scope of information acquisition and understand the entities better.Linked Data uses URIs to represent things and establish linking between datasets through URIs.The same entities with different identifiers or URIs distributed in different datasets are connected to each other by owl:sameAs.In the Web of Data,an increasing number of owl:sameAs statements have been published to support merging distributed descriptions of equivalent RDF resources.In order to better understand and use owl:sameAs in Linked Data,it is useful to study how owl:sameAs is actually deployed,which has implications for how data should be consumed.To the best of our knowledge,the study on equivalent entity network extracted from the real world Web of data is very little.This pape wants to find the deployment and application of owl: sameAs in the web of data.Based on a large number of owl: sameAs links extracted from the BTC 2014 dataset in the real web of data,this paper analyzes the structure of the sample equivalent entity network,the domain names and instance types.For the equivalent entity network,random failure of most nodes will not affect its connectivity,but the failure of a few important nodes will have a certain impact on the connectivity of the network.In the equivalent entity network,the entity type that establishes the most owl:sameAs links is the person name,followed by the place name,the medicine name,the organization name,and the movie.On the basis of detailed analysis of entity identity,this paper sums up the attribute correspondences of instance pairs in the equivalent entity network,discusses the problem of instances of different attribute values under the same attribute in the equivalent entity network and finally discusses the dynamic property and maintenance of sameAs links.