LCC Tools for Question Answering

Abstract

The increased complexity of the TREC QA questions requires advancedtext processing tools that rely on natural language processing andknowledge reasoning. This paper presents the suite of tools thataccount for the performance of the PowerAnswer question answeringsystem. It is shown how questions, answers and world knowledge aretransformed first in logic representation, followed by a systematicand rigorous logic proof that validly answers questions posed tothe QA system. At TREC QA 2002, PowerAnswer obtained a confidence-weightedscore of 0.856, answering correctly 415 out of 500 questions.