Automatic Generation of Java/SQL Based Inference Engines from RDF Schema and RuleML

This paper describes two approaches for automatically converting RDF Schema and RuleML sources into an inference engine and storage repository. Rather than using traditional inference systems, our solution bases on mainstream technologies like Java and relational database systems. While this necessarily imposes some restrictions, the ease of integration into an existing IT landscape is a major advantage. We present the conversion tools and their limitations. Furthermore, an extension to RuleML is proposed, that allows Java-enabled reaction rules, where calls to Java libraries can be performed upon a rule firing. This requires hosts to be Java-enabled when rules and code are moved across the web. However, the solution allows for great engineering flexibility.