Host: The Japanese Society for Artificial Intelligence
Name : The 26th Annual Conference of the Japanese Society for Artificial Intelligence, 2012
Number : 26
Location : [in Japanese]
Date : June 12, 2012 - June 15, 2012
The Linked Open Data (LOD) includes over 31 billion Resource Description Framework (RDF) triples, but only interlinked by around 504 million RDF links. Linking related or identical instances in the LOD can help semantic web application developers easily collect data from various data sets. However, because of the heterogeneity of ontology schema in the LOD, it is difficult for them to query the data sets without manually learning ontology schema. An automatic ontology schema integration method can help us to detect and integrate important ontology schema for linking related data. Since the links between related resources construct a linked network, we can detect linked graph pattern from the linked data. By analyzing the graph patterns, we can discover related predicates and the core predicates that are used for linking related instances. Our approach performs ontology similarity matching method on the linked graph patterns to identify related predicates from different ontology schema. Using the automatically integrated ontology schema, semantic web application developers can easily understand the links between instances and effectively query on the linked data sets.
