Answering complex, list and context questions with LCC's Question-Answering Server
Sanda M. Harabagiu, DAN I. MOLDOVAN, MARIUS A. PAŞCA, Mihai Surdeanu, Rada F. Mihalcea, Corina R Girju, Vasile Rus, Finley Lăcătuşu, Paul Morǎrescu, Răzvan Bunescu · University of North Texas Digital Library (University of North Texas) · 2001
This paper presents the architecture of the Question-Answering Server (QAS) developed at the Language Computer Corporation (LCC) and used in the TREC-10 evaluations. LCC's QAS extracts answers for (a) factual questions of variable degree of difficulty; (b) questions that expect lists of answers; and (c) questions posed in the context of previous questions and answers. One of the major novelties is the implementation of bridging inference mechanisms that guide the search for answers to complex questions. Additionally, LCC's QAS encodes an efficient way of modeling context via reference resolution. In TREC-10, this system generated an RAR, of 0.58 on the main task and 0.78 on the context task.