Bridging data mining and semantic web

Abstract

Nowadays Semantic Web is widely adopted standard of knowledge representation. Hence, knowledge engineers are applying sophisticated methods to capture, discover and represent knowledge in Semantic Web form. Studies show that, to represent knowledge in Semantic Web standard, data mining techniques such as Decision Trees, Association Rules, etc., play an important role. These techniques are implemented in publicly available Data Mining tools. These tools represent knowledge discovered in human readable format and some tools use Predictive Model Markup language (PMML).

PMML is an XML based model for data mining model representation that fails to address the representation of the semantics of the discovered knowledge. Hence, this thesis tries to research and give solutions to translate PMML to Semantic Web Rule Language (SWRL) format using Semantic Web technologies and data mining to cover the semantic gap in PMML.