Title: Resource propagation algorithm considering predicates to complement knowledge bases in linked data

Authors: Toshitaka Maki; Kazuki Takahashi; Toshihiko Wakahara; Akihisa Kodate; Noboru Sonehara

Addresses: Intelligent Information System Engineering, Graduate School, Fukuoka Institute of Technology, 3-30-1 Wajirohigashi, Higashi-ku, Fukuoka-shi, Fukuoka, Japan ' Communication and Information Engineering, Graduate School, Fukuoka Institute of Technology, 3-30-1 Wajirohigashi, Higashi-ku, Fukuoka-shi, Fukuoka, Japan ' Intelligent Information System Engineering, Graduate School, Fukuoka Institute of Technology, 3-30-1 Wajirohigashi, Higashi-ku, Fukuoka-shi, Fukuoka, Japan ' College of Policy Studies, Tsuda University, 1-18-24 Sendagaya, Shibuya-ku, Tokyo, Japan ' College of Policy Studies, Tsuda University, 1-18-24 Sendagaya, Shibuya-ku, Tokyo, Japan

Abstract: Linked data are directed graph data with descriptive labels using uniform resource identifiers (URI) and based on resource description framework (RDF). Linked data which are published as open data on the web are called linked open data (LOD). LOD is popular as a technology for constructing the semantic web because it can create a knowledge base by linking various resources. However, a large amount of the linked data does not have sufficient links since resources defined by URI type are scarce. Therefore, this paper presents a new resource propagation algorithm (RPA) which predicts links between resources in the linked data, and complements the knowledge bases by considering predicates in the RDF structure. The presented experiment demonstrates that the RPA was able to complement semantic links between resources considering each predicate.

Keywords: open data; linked data; RDF; graph; predicate; link prediction.

DOI: 10.1504/IJSSC.2018.094495

International Journal of Space-Based and Situated Computing, 2018 Vol.8 No.2, pp.115 - 121

Received: 17 Oct 2017
Accepted: 14 May 2018
Published online: 03 Sep 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article