The impact of intent selection on diversified search evaluation

Tetsuya Sakai, Zhicheng Dou, Charles L A Clarke

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

To construct a diversified search test collection, a set of possible subtopics (or intents) needs to be determined for each topic, in one way or another, and per-intent relevance assessments need to be obtained. In the TREC Web Track Diversity Task, subtopics are manually developed at NIST, based on results of automatic click log analysis; in the NTCIR INTENT Task, intents are determined by manually clustering "subtopics strings" returned by participating systems. In this study, we address the following research question: Does the choice of intents for a test collection affect relative performances of diversified search systems? To this end, we use the TREC 2012 Web Track Diversity Task data and the NTCIR-10 INTENT-2 Task data, which share a set of 50 topics but have different intent sets. Our initial results suggest that the choice of intents may affect relative performances, and that this choice may be far more important than how many intents are selected for each topic.

Original languageEnglish
Title of host publicationSIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval
Pages921-924
Number of pages4
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event36th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2013 - Dublin
Duration: 2013 Jul 282013 Aug 1

Other

Other36th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2013
CityDublin
Period13/7/2813/8/1

Keywords

  • Diversity
  • Evaluation
  • Intents
  • Subtopics
  • Test collections

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Information Systems

Cite this

Sakai, T., Dou, Z., & Clarke, C. L. A. (2013). The impact of intent selection on diversified search evaluation. In SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 921-924) https://doi.org/10.1145/2484028.2484105

The impact of intent selection on diversified search evaluation. / Sakai, Tetsuya; Dou, Zhicheng; Clarke, Charles L A.

SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2013. p. 921-924.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Sakai, T, Dou, Z & Clarke, CLA 2013, The impact of intent selection on diversified search evaluation. in SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval. pp. 921-924, 36th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2013, Dublin, 13/7/28. https://doi.org/10.1145/2484028.2484105
Sakai T, Dou Z, Clarke CLA. The impact of intent selection on diversified search evaluation. In SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2013. p. 921-924 https://doi.org/10.1145/2484028.2484105
Sakai, Tetsuya ; Dou, Zhicheng ; Clarke, Charles L A. / The impact of intent selection on diversified search evaluation. SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2013. pp. 921-924
@inproceedings{a2453f4010cc470c99b0c8fdb7e466ec,
title = "The impact of intent selection on diversified search evaluation",
abstract = "To construct a diversified search test collection, a set of possible subtopics (or intents) needs to be determined for each topic, in one way or another, and per-intent relevance assessments need to be obtained. In the TREC Web Track Diversity Task, subtopics are manually developed at NIST, based on results of automatic click log analysis; in the NTCIR INTENT Task, intents are determined by manually clustering {"}subtopics strings{"} returned by participating systems. In this study, we address the following research question: Does the choice of intents for a test collection affect relative performances of diversified search systems? To this end, we use the TREC 2012 Web Track Diversity Task data and the NTCIR-10 INTENT-2 Task data, which share a set of 50 topics but have different intent sets. Our initial results suggest that the choice of intents may affect relative performances, and that this choice may be far more important than how many intents are selected for each topic.",
keywords = "Diversity, Evaluation, Intents, Subtopics, Test collections",
author = "Tetsuya Sakai and Zhicheng Dou and Clarke, {Charles L A}",
year = "2013",
doi = "10.1145/2484028.2484105",
language = "English",
isbn = "9781450320344",
pages = "921--924",
booktitle = "SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval",

}

TY - GEN

T1 - The impact of intent selection on diversified search evaluation

AU - Sakai, Tetsuya

AU - Dou, Zhicheng

AU - Clarke, Charles L A

PY - 2013

Y1 - 2013

N2 - To construct a diversified search test collection, a set of possible subtopics (or intents) needs to be determined for each topic, in one way or another, and per-intent relevance assessments need to be obtained. In the TREC Web Track Diversity Task, subtopics are manually developed at NIST, based on results of automatic click log analysis; in the NTCIR INTENT Task, intents are determined by manually clustering "subtopics strings" returned by participating systems. In this study, we address the following research question: Does the choice of intents for a test collection affect relative performances of diversified search systems? To this end, we use the TREC 2012 Web Track Diversity Task data and the NTCIR-10 INTENT-2 Task data, which share a set of 50 topics but have different intent sets. Our initial results suggest that the choice of intents may affect relative performances, and that this choice may be far more important than how many intents are selected for each topic.

AB - To construct a diversified search test collection, a set of possible subtopics (or intents) needs to be determined for each topic, in one way or another, and per-intent relevance assessments need to be obtained. In the TREC Web Track Diversity Task, subtopics are manually developed at NIST, based on results of automatic click log analysis; in the NTCIR INTENT Task, intents are determined by manually clustering "subtopics strings" returned by participating systems. In this study, we address the following research question: Does the choice of intents for a test collection affect relative performances of diversified search systems? To this end, we use the TREC 2012 Web Track Diversity Task data and the NTCIR-10 INTENT-2 Task data, which share a set of 50 topics but have different intent sets. Our initial results suggest that the choice of intents may affect relative performances, and that this choice may be far more important than how many intents are selected for each topic.

KW - Diversity

KW - Evaluation

KW - Intents

KW - Subtopics

KW - Test collections

UR - http://www.scopus.com/inward/record.url?scp=84883100429&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883100429&partnerID=8YFLogxK

U2 - 10.1145/2484028.2484105

DO - 10.1145/2484028.2484105

M3 - Conference contribution

SN - 9781450320344

SP - 921

EP - 924

BT - SIGIR 2013 - Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval

ER -