Literal Node Matching Based on Image Features toward Linked Data Integration

Abstract

Linked Open Data (LOD) has a graph structure in which nodes are represented by Uniform Resource Identifiers (URIs), and thus LOD sets are connected and searched through different domains. In fact, however, 5% of the values are literal (string without URI) even in DBpedia, which is a de facto hub of LOD. Since the literal becomes a terminal node, and we need to rely on regular expression matching, we cannot trace the links in the LOD graphs during searches. Therefore, this paper proposes a method of identifying and aggregating literal nodes that have the same meaning in order to facilitate cross-domain search through links in LOD. The novelty of our method is that part of the LOD graph structure is regarded as a block image, and then image features of LOD are extracted. In experiments, we created about 30,000 literal pairs from a Japanese music category of DBpedia Japanese and Freebase, and confirmed that the proposed method correctly determines literal identity with F-measure of 99%.