Information Extraction from TextErik Tjong Kim Sang -
8 May 2007 (abstract)Abstract: Information extraction deals with mining text for factual relations between entities. Traditionally this task involved employing manually defined text patterns which were designed with a particular relation in mind. Recently, research focus has shifted to extraction patterns which are learned from a combination of predefined sets of examples of relation elements and text. In this talk, I will describe our current work on extracting relational information from text. The work is being applied for building answer tables for question answering systems as well as extending general linguistic resources such as the Dutch part of EuroWordNet.
