Linked data approach in accessing geospatial big data

Abstract

Today, linked data is frequently associated with Geographic Information Systems (GIS) as its technology stack is utilized in alleviating geospatial data integration issue. Geospatial data have become ubiquitous as they have emerged everywhere and these data can be geo-referenced. One of the types of georeferenced data that is lacking in Malaysia is the insufficient availability of Malaysian oceanographic data. It is a great relief to know that most of the earth observation agencies have granted access into their data obtained from satellite altimetry. Consequently, the exponential growth of geospatial data as well as its complexity and diversity has led to big data problem and caused information sharing and exchange on the web becoming more complicated. To resolve this issue, linked data should be used in handling geospatial big data. Linked data is one of the best practices for exposing, sharing, publishing and connecting the structured data on the web. This study explored linked data as an approach to provide access to the Malaysian physical oceanography datasets on the web, which would allow the data to be standardized in a machine-readable format. The research reviewed the existing software tools used in publishing linked data, identified an appropriate software tool to generate Resource Description Framework (RDF) presenting geographical data and built a physical oceanography data website based on linked data principles. Initially, document analysis was conducted to review the existing linked data tools that have been used for geospatial data. Various scholarly articles, journals, tutorials and web pages were used as references to investigate the use of linked data tools. Based on the review, five software tools, namely Geometry2RDF, TripleGeo, Datalift, OpenLink Virtuoso and KARMA were identified as the appropriate tools to generate the RDF. Each of this software tool has its own capabilities and functionalities. Next, the tools were compared with one another based on literature review to get the best possible tool that can manage georeferenced oceanographic data. After the comparison, this study identified the best software tool to transform the shapefile into the RDF format was Datalift. Finally, a web-based information system was built to publish the linked data to data interlinking and sharing by web users. In conclusion, this study has introduced an alternative way to publish and access geospatial data, particularly related to physical oceanography datasets using linked data principles. Using such an approach would facilitate stakeholders and unveil information within the big data, thus enriching the discovery of geospatial information on the web.