Towards semantic data mining

Abstract

Incorporating domain knowledge is one of the most challenging problems in data mining. The Semantic Web technologies are promising to offer solutions to formally capture and efficiently use the domain knowledge. We call data mining technologies powered by the Semantic Web, capable of systematically incorporating domain knowledge, the semantic data mining. In this paper, we identify the importance of semantic annotation-a crucial step towards realizing semantic data mining by bringing meaning to data, and propose a learning-based semantic search algorithm for annotating (semi-) structured data.