Ontology Schema-specific Rule Materialization

The reasoning should tackle big data issues like other domains as the size of ontology grows bigger and bigger. Especially, rule-based reasoning should overcome the following challenges: duplicate elimination and rule matching efficiency. To deal with these challenges, we introduce a new rulebased reasoning method which materializes each generic instance rule into several schema-specific instance rules and combines with Hadoop framework to deal with billions of triples. The experiment shows the materialization remarkably improves the efficiency of rule-based reasoning by reducing the amount of required memory and making it linear to the data size.