Department of Computer Science and Technology

Technical reports

Steps towards natural language to data language translation using general semantic information

B.K. Boguraev, K. Spärck Jones

March 1982, 8 pages

Abstract

The aim of the work reported here is to maximise the use of general semantic information in an AI task processor, specifically in a system front end for converting natural language questions into formal database queries. The paper describes the translation component of such a front end, which is designed to work from the question meaning representation produced by a language analyser exploiting only general semantics and syntax, to a formal query relying on database-specific semantics and syntax. Translation is effected in three steps, and the paper suggests that the rich and explicit meaning representations using semantic primitives produced for input sentences by the analyser constitute a natural and effective base for further processing.

Full text

PDF (0.5 MB)

BibTeX record

@TechReport{UCAM-CL-TR-24,
  author =	 {Boguraev, B.K. and Sp{\"a}rck Jones, K.},
  title = 	 {{Steps towards natural language to data language
         	   translation using general semantic information}},
  year = 	 1982,
  month = 	 mar,
  url = 	 {http://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-24.pdf},
  institution =  {University of Cambridge, Computer Laboratory},
  number = 	 {UCAM-CL-TR-24}
}