Publications
Authors:
  • Greg Grefenstette
Citation:
ASIS Annual Meeting, Pittsburgh, PA October 26-29, 1998.
Abstract:
As information becomes globally accessible, when foreign language documents are returned as a result of a
query, they can either be regarded as a nuisance to be discarded, or as an unmined source of potential
answers to the original query. Taking the second view is the new research area called Cross Language
Information Retrieval. This paper will discuss the problems and current research techniques being explored in
this area for finding documents in languages other than the original query. We will talk about language
identification methods (n-grams and short word techniques), techniques for automatically generating queries in
other languages (stemming and morphological analysis; dictionary-based translation, corpus-based translation,
machine-translation based; query conflation), and retrieving and merging query results (weighting schemes)
Year:
1998
Report number:
1998/303