论文信息 - Less is more: Eliminating index terms from subordinate clauses

Less is more: Eliminating index terms from subordinate clauses

We perform a linguistic analysis of documents during indexing for information retrieval. By eliminating index terms that occur only in subordinate clauses, index size is reduced by approximately 30% without adversely affecting precision or recall. These results hold for two corpora: a sample of the world wide web and an electronic encyclopedia.

William B. Dolan | Simon Corston-Oliver

[1] William C. Mann,et al. Rhetorical Structure Theory: Toward a functional theory of text organization , 1988 .

[2] Boris Katz,et al. Annotating the World Wide Web using Natural Language , 1997, RIAO.

[3] Boris Katz,et al. Annotating the World Wide Web , 2001 .

[4] Avi Arampatzis,et al. Phrase-based Information Retrieval , 1998 .

[5] Tomek Strzalkowski. Natural Language Information Retrieval , 1995, Inf. Process. Manag..

[6] William C. Mann,et al. Relational propositions in discourse , 1986 .

[7] Sandra A. Thompson,et al. The structure of discourse and ‘subordination’ , 1987 .

[8] Tomek Strzalkowski,et al. Natural Language Information Retrieval: TREC-8 Report , 1994, TREC.

[9] S. Corston-Oliver,et al. Computing representations of the structure of written discourse , 1998 .

[10] Gregory Grefenstette,et al. Cross-Language Information Retrieval , 1998, The Springer International Series on Information Retrieval.

[11] Avi Arampatzis,et al. Phase-Based Information Retrieval , 1998, Inf. Process. Manag..

[12] Joel L Fagan,et al. Experiments in Automatic Phrase Indexing For Document Retrieval: A Comparison of Syntactic and Non-Syntactic Methods , 1987 .