DMCA
Xerox TREC-8 Question Answering Track Report (0)
| Venue: | In Voorhees and Harman [21 |
| Citations: | 19 - 0 self |
BibTeX
@INPROCEEDINGS{Hull_xeroxtrec-8,
author = {David A. Hull},
title = {Xerox TREC-8 Question Answering Track Report},
booktitle = {In Voorhees and Harman [21},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
This report describes the Xerox work on the TREC-8 Question Answering Track. We linked together a few basic NLP components (a question parser, a sentence boundary identifier, and a proper noun tagger) with a sentence scoring function and an answer presentation function built specifically for the TREC Q&A task. Our system found the correct 50-byte answer (in the top 5 responses) to 45% of the questions, a quite respectable performance, but with considerable room for improvement. Based on the failure analysis presented in this paper, we can conclude that the system would benefit from having access to a broad range of other NLP technologies, including robust parsing and coreference analysis, or some good heuristic approximations thereof. The system also has a clear need for some semantic resources to help with certain difficult problems, such as finding answers that match the semantic class X in What X? questions. 1 Introduction Natural language processing (NLP) techniques have...
Keyphrases
broad range respectable performance correct 50-byte answer xerox work robust parsing good heuristic approximation thereof basic nlp component trec-8 question answering track sentence boundary identifier trec task coreference analysis answer presentation function failure analysis nlp technology proper noun tagger introduction natural language processing clear need semantic class certain difficult problem question parser semantic resource considerable room
