Converting Open Data to Linked Data -Automated Vocabulary Unication and Linking-

Abstract

In converting open data which are published in form of CSV("Comma Separeted Values") or spread sheet into Linked Data, especially to use the data in combination with other Linked Data, they should be written with unified vocaburaly and include links to other data sets. We report technologies to support converting open data into Linked Data by automating vocabulary unification and linking to other linked data. For vocaburaly unification, we show vocaburaly suggestion technology which searches existing LOD("Linked Open Data") and infers a suitable vocaburaly set for a given CSV. For automatic linking, we show identity inference technology which calculates degree of simularity between entities in CSV and LOD. These technologies help people converting there data into Linked Data, but we notice that we need to build a mechanism to incorporate human knowledge in order to correct errors.