| Douglas W. Oard. Alignment of Spanish and English TREC topic descriptions. In D. K. Harman, editor, The Fifth Text REtrieval Conference (TREC-5). NIST, November 1996. http://www.glue.umd.edu/¸oard/research.html. |
....Clearly, as documents become less and less comparable, using them for CLIR becomes more and more difficult. For convenience, we assume that comparable corpora are not parallel in the following discussions. Dumais, Landauer, and Littman at Bellcore [14, 9] and Oard at the Univerity of Maryland [17, 9] have used a matrix reduction technique called Latent Semantic Indexing (LSI) to extract language independent term and document representations from a parallel corpus. Essentialy, LSI applies a singular value decomposition to the large, sparse term document cooccurrence matrix (including terms ....
Douglas W. Oard. Alignment of spanish and english TREC topic descriptions. In The 5th Text Retrieval Conference (TREC-5), 1997. To appear.
No context found.
Douglas W. Oard. Alignment of Spanish and English TREC topic descriptions. In D. K. Harman, editor, The Fifth Text REtrieval Conference (TREC-5). NIST, November 1996. http://www.glue.umd.edu/¸oard/research.html.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC