A maximum entropy framework that integrates word dependencies and grammatical relations for reading comprehension

Authors:
Kui Xu;Helen Meng;Fuliang Weng
Affiliations:
The Chinese University of Hong Kong, Hong Kong SAR, China and Robert Bosch Corp., Palo Alto, CA;The Chinese University of Hong Kong, Hong Kong SAR, China;Robert Bosch Corp., Palo Alto, CA
Venue:
NAACL-Short '06 Proceedings of the Human Language Technology Conference of the NAACL, Companion Volume: Short Papers
Year:
2006

Citing 8
Cited 0

Head-driven statistical models for natural language parsing

Head-driven statistical models for natural language parsing
Analyses for elucidating current question answering technology

Natural Language Engineering
Deep Read: a reading comprehension system

ACL '99 Proceedings of the 37th annual meeting of the Association for Computational Linguistics on Computational Linguistics
Question answering using maximum entropy components

NAACL '01 Proceedings of the second meeting of the North American Chapter of the Association for Computational Linguistics on Language technologies
Reading comprehension programs in a statistical-language-processing class

ANLP/NAACL-ReadingComp '00 Proceedings of the 2000 ANLP/NAACL Workshop on Reading comprehension tests as evaluation for computer-based language understanding sytems - Volume 6
A rule-based question answering system for reading comprehension tests

ANLP/NAACL-ReadingComp '00 Proceedings of the 2000 ANLP/NAACL Workshop on Reading comprehension tests as evaluation for computer-based language understanding sytems - Volume 6
A fast algorithm for feature selection in conditional maximum entropy modeling

EMNLP '03 Proceedings of the 2003 conference on Empirical methods in natural language processing
The use of metadata, web-derived answer patterns and passage context to improve reading comprehension performance

HLT '05 Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Language Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Automatic reading comprehension (RC) systems can analyze a given passage and generate/extract answers in response to questions about the passage. The RC passages are often constrained in their lengths and the target answer sentence usually occurs very few times. In order to generate/extract a specific precise answer, this paper proposes the integration of two types of "deep" linguistic features, namely word dependencies and grammatical relations, in a maximum entropy (ME) framework to handle the RC task. The proposed approach achieves 44.7% and 73.2% HumSent accuracy on the Remedia and ChungHwa corpora respectively. This result is competitive with other results reported thus far.