Extracting Common Sense Knowledge from Wikipedia
Sangweon Suh, Harry Halpin, Ewan Klein · 2006
Much of the natural language text found on the web contains various kinds of generic or “common sense ” knowledge, and this information has long been recognized by artificial intelligence as an important supplement to more formal approaches to building Semantic Web knowledge bases. Consequently, we are exploring the possibility of automatically identifying “common sense” statements from unrestricted natural language text and mapping them to RDF. Our hypothesis is that common sense knowledge is often expressed in the form of generic statements such as Coffee is a popular beverage, and thus our work has focussed on the challenge of automatically identifying generic statements. We have been using the Wikipedia xml corpus as a rich source of common sense knowledge. For evaluation, we have been using the existing annotation of generic entities and relations in the ace 2005 corpus.