Consolidation of Linked Data Resources upon Heterogeneous Schemas

摘要

Linked data resources have influential roles in conducting the future of semantic web. They are growing more and more, and the amount of published data is increasing at a fast pace. It causes some new concerns arise in the context of semantic web. One of the most important issues is the large amount of data that is produced as identical entities in heterogeneous data sources by different providers. This is a barrier to intelligent applications or agents that are going to utilize linked data resources. It prevents us from utilizing the potential capacity of web of data. Linked data resources are valuable when we could exploit them altogether. Therefore, we could obviously perceive the importance of linked data integration. In this paper, we propose a new approach for linked data consolidation. It helps us to have a consolidation process even between resources with heterogeneous schemas. In this approach, we are going to find more identical instances locally. This means that we direct our instance coreference resolution around the two instances which are certainly identical. The neighbors of two similar instances are a good source for our approach to proceed. In addition, these neighbors are beneficial for estimating some similarities between concepts of two heterogeneous schemas.