On building a more efficient grammar by exploiting types
Natural Language Engineering
A bag of useful techniques for efficient and robust parsing
ACL '99 Proceedings of the 37th annual meeting of the Association for Computational Linguistics on Computational Linguistics
A parsing: fast exact Viterbi parse selection
NAACL '03 Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology - Volume 1
COLING-GEE '02 Proceedings of the 2002 workshop on Grammar engineering and evaluation - Volume 15
Deep processing for a portable natural language interface to databases
Deep processing for a portable natural language interface to databases
| Hi-index | 0.00 |
This paper reports on one aspect of Locutus, a natural language interface to databases (NLIDB) which uses the output of a high-precision broad-coverage grammar to build semantic representations and ultimately SQL queries. Rather than selecting just a subset of the parses provided by the grammar to use in further processing, Locutus uses all of them. If the meaning of a parse does not conform to the semantic domain of the database, no query is built for it. Thus, intended parses are chosen extrinsically. The parser gives an average of 3.01 parses to the sentences in the Geoquery250 corpus. Locutus generates an average of 1.02 queries per sentence for this corpus, all of them correct.