Structure analysis and a schema definition method for creating Linked Open Data from complex information resources

Abstract

Linked Open Data (LOD) have attracted much attention as a method for publishing data using Web technology. However, since many information resources on the Web are organized for human reading, they are not well structured for LOD. In this study, we investigate how to create LOD from complex information resources with different types of elements, achieving interoperability without any information loss. We found that by analyzing the structure and vocabulary based on the description of the information resource, it is possible to structure information efficiently to create LOD. Furthermore, we propose an approach for applying the appropriate LOD structure that compensates lack of information with linking other resources when using information resources designed for human reading.