[comp.ai] Text Analysis for Text Retrieval

finin@prc.unisys.com (Tim Finin) (03/30/89)

				 AI SEMINAR
			UNISYS PAOLI RESEARCH CENTER
				      
				      
		      Text Analysis for Text Retrieval
				      
			       David D. Lewis
		      Information Retrieval Laboratory
		   University of Massachusetts at Amherst

The performance of traditional text retrieval methods has plateaued in recent
years, sparking a renewed interest in applications of natural language
processing (NLP) to information retrieval (IR). The ADRENAL (Augmented
Document REtrieval using NAtural Language processing) system, under
construction at U Mass, makes use of syntactic and semantic processing, as
well as plausible inference techniques, to construct a rich, though rather
general and errorful, semantic representation of queries and documents for
use in retrieval. I will focus in this talk on the problems of applying NLP
to large collections of real-world text, paying particular attention to the
lexical analysis and inference phases. Preliminary results from hand
simulations of the system, and data on the operational syntactic parser and
inference components, will be presented. Two algorithms which have been
implemented for making use of NLP-produced representations in text retrieval
will be described. One is based on the probabilistic retrieval model from IR,
while the other is an incremental graph matching and inference algorithm
derived from AI work on knowledge representation.
				      
				      
			  11:00am Monday, April 17
			     BIC Conference Room
			Unisys Paoli Research Center
			 Route 252 and Central Ave.
			       Paoli PA 19311
				      
      -- non-Unisys visitors who are interested in attending should --
      --   send email to finin@prc.unisys.com or call 215-648-7446  --
-- 
 Tim Finin			finin@prc.unisys.com (internet)
 Unisys Paoli Research Center	..!{psuvax1,sdcrdcf,cbmvax}!burdvax!finin (uucp)
 PO Box 517			215-648-7446 (office), 215-386-1749 (home),
 Paoli, PA 19301		215-648-7412 (fax)