University of Sheffield TREC-8 Q & A System
Kevin Humphreys, Robert Gaizauskas, Mark Hepple, Mark Sanderson · Text REtrieval Conference · 1999
The system entered by the University of Sheffield in the question answering track of TREC-8 is the result of coupling two existing technologies - information retrieval (IR) and information extraction (IE). In essence the approach is this: the IR system treats the question as a query and returns a set of top ranked documents or passages; the IE system uses NLP techniques to parse the question, analyse the top ranked documents or passages returned by the IR system, and instantiate a query variable in the semantic representation of the question against the semantic representation of the analysed documents or passages. Thus, while the IE system by no means attempts “full text understanding, this approach is a relatively deep approach which attempts to work with meaning representations. Since the information retrieval systems we used were not our own (AT&T and UMass) and were used more or less “off the shelf, this paper concentrates on describing the modifications made to our existing information extraction system to allow it to participate in the Q & A task.