Shadow Document Methods of Results Merging

S. Wu, F. Crestani

Research output: Contribution to conferencePaper

17 Citations (Scopus)

Abstract

In distributed information retrieval systems, document overlaps occur frequently across results from different databases. This is especially the case for meta-search engines which merge results from several general-purpose web search engines. This paper addresses the problem of merging results which contain overlaps in order to achieve better performance. Several algorithms for merging results are proposed, which take advantage of the use of duplicate documents in two ways: one correlates scores from different results; the other regards duplicates as increasing evidence of being relevant to the given query. A variety of experiments have demonstrated that these methods are effective.
LanguageEnglish
Pages1067-1072
Number of pages5
Publication statusPublished - 17 Mar 2004
EventProceedings of ACM SAC 2004 - Nicosia, Cyprus
Duration: 14 Mar 200417 Mar 2004

Conference

ConferenceProceedings of ACM SAC 2004
CityNicosia, Cyprus
Period14/03/0417/03/04

Fingerprint

Search engines
Merging
Information retrieval systems
Experiments

Keywords

  • merging results
  • algorithms
  • meta-search engines
  • information retrieval

Cite this

Wu, S., & Crestani, F. (2004). Shadow Document Methods of Results Merging. 1067-1072. Paper presented at Proceedings of ACM SAC 2004, Nicosia, Cyprus, .
Wu, S. ; Crestani, F. / Shadow Document Methods of Results Merging. Paper presented at Proceedings of ACM SAC 2004, Nicosia, Cyprus, .5 p.
@conference{3198404bdbbb4cb38b37ec69da6b0670,
title = "Shadow Document Methods of Results Merging",
abstract = "In distributed information retrieval systems, document overlaps occur frequently across results from different databases. This is especially the case for meta-search engines which merge results from several general-purpose web search engines. This paper addresses the problem of merging results which contain overlaps in order to achieve better performance. Several algorithms for merging results are proposed, which take advantage of the use of duplicate documents in two ways: one correlates scores from different results; the other regards duplicates as increasing evidence of being relevant to the given query. A variety of experiments have demonstrated that these methods are effective.",
keywords = "merging results, algorithms, meta-search engines, information retrieval",
author = "S. Wu and F. Crestani",
year = "2004",
month = "3",
day = "17",
language = "English",
pages = "1067--1072",
note = "Proceedings of ACM SAC 2004 ; Conference date: 14-03-2004 Through 17-03-2004",

}

Wu, S & Crestani, F 2004, 'Shadow Document Methods of Results Merging' Paper presented at Proceedings of ACM SAC 2004, Nicosia, Cyprus, 14/03/04 - 17/03/04, pp. 1067-1072.

Shadow Document Methods of Results Merging. / Wu, S.; Crestani, F.

2004. 1067-1072 Paper presented at Proceedings of ACM SAC 2004, Nicosia, Cyprus, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Shadow Document Methods of Results Merging

AU - Wu, S.

AU - Crestani, F.

PY - 2004/3/17

Y1 - 2004/3/17

N2 - In distributed information retrieval systems, document overlaps occur frequently across results from different databases. This is especially the case for meta-search engines which merge results from several general-purpose web search engines. This paper addresses the problem of merging results which contain overlaps in order to achieve better performance. Several algorithms for merging results are proposed, which take advantage of the use of duplicate documents in two ways: one correlates scores from different results; the other regards duplicates as increasing evidence of being relevant to the given query. A variety of experiments have demonstrated that these methods are effective.

AB - In distributed information retrieval systems, document overlaps occur frequently across results from different databases. This is especially the case for meta-search engines which merge results from several general-purpose web search engines. This paper addresses the problem of merging results which contain overlaps in order to achieve better performance. Several algorithms for merging results are proposed, which take advantage of the use of duplicate documents in two ways: one correlates scores from different results; the other regards duplicates as increasing evidence of being relevant to the given query. A variety of experiments have demonstrated that these methods are effective.

KW - merging results

KW - algorithms

KW - meta-search engines

KW - information retrieval

UR - http://www.acm.org/conferences/sac/sac2004/

UR - http://dx.doi.org/10.1145/967900.968117

M3 - Paper

SP - 1067

EP - 1072

ER -

Wu S, Crestani F. Shadow Document Methods of Results Merging. 2004. Paper presented at Proceedings of ACM SAC 2004, Nicosia, Cyprus, .