Exploring Scientific Publication and Cross Domain Linked Dataset for Similarity - A Case Study

Abstract

Linking Open Data Project played a vital role in the realization of structured data at World Wide Web stage by methodically demonstrating the importance of machine understandable data for information linking. It has succeeded in bringing up heap of Linked Open Data ranging from geographic to cross-domain datasets which provide huge opportunities for knowledge discovery and mashup application development. Scientific publication datasets are one of main sources in steering today's research work and has a big share in Linked Data Cloud repository. Besides to it, crossdomain linked data datasets e.g. DBpedia, FreeBase etc. has a huge crowd-sourced open knowledge which proved as good resource for content enrichment and interlinking. Noticing the offered added values of scientific publications and cross-domain datasets it will be great to know; what these datasets has to offer each other in Linked Data settings. We are of a view; if these datasets are interlinked can offer adequate information for enrichment of publication related resources i.e. authors and publications. In addition, this will also help to aggregate information of author in a profile which is currently scattered over different linked data resources. However, currently finding and interlinking with appropriate data is still a challenge in Linked Data Cloud. In this paper we presented a case study by interlinking author from scientific publication dataset (DBLP) with person’s record of crossdomain dataset (DBpedia). Moreover, we have investigated to find how much author information is there in DBpedia for indexed DBLP scientific authors and has validated our assumption that meaningful data is present between these datasets.