Estimating intent types for search result diversification

Kosetsu Tsukuda, Tetsuya Sakai, Zhicheng Dou, Katsumi Tanaka

    研究成果: Conference contribution

    4 引用 (Scopus)

    抄録

    Given an ambiguous or underspecified query, search result diversification aims at accommodating different user intents within a single Search Engine Result Page (SERP). While automatic identification of different intents for a given query is a crucial step for result diversification, also important is the estimation of intent types (informational vs. navigational). If it is possible to distinguish between informational and navigational intents, search engines can aim to return one best URL for each navigational intent, while allocating more space to the informational intents within the SERP. In light of the observations, we propose a new framework for search result diversification that is intent importance-aware and type-aware. Our experiments using the NTCIR-9 INTENT Japanese Subtopic Mining and Document Ranking test collections show that: (a) our intent type estimation method for Japanese achieves 64.4% accuracy; and (b) our proposed diversification method achieves 0.6373 in D#-nDCG and 0.5898 in DIN#-nDCG over 56 topics, which are statistically significant gains over the top performers of the NTCIR-9 INTENT Japanese Document Ranking runs. Moreover, our relevance oriented model significantly outperforms our diversity oriented model and the original model by Dou et al..

    元の言語English
    ホスト出版物のタイトルLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    ページ25-37
    ページ数13
    8281 LNCS
    DOI
    出版物ステータスPublished - 2013
    イベント9th Asia Information Retrieval Societies Conference on Information Retrieval Technology, AIRS 2013 - Singapore
    継続期間: 2013 12 92013 12 11

    出版物シリーズ

    名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    8281 LNCS
    ISSN(印刷物)03029743
    ISSN(電子版)16113349

    Other

    Other9th Asia Information Retrieval Societies Conference on Information Retrieval Technology, AIRS 2013
    Singapore
    期間13/12/913/12/11

    Fingerprint

    Diversification
    Search engines
    Search Engine
    Ranking
    Websites
    Query
    Ambiguous
    Mining
    Model
    Experiments
    Experiment

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    これを引用

    Tsukuda, K., Sakai, T., Dou, Z., & Tanaka, K. (2013). Estimating intent types for search result diversification. : Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (巻 8281 LNCS, pp. 25-37). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 巻数 8281 LNCS). https://doi.org/10.1007/978-3-642-45068-6_3

    Estimating intent types for search result diversification. / Tsukuda, Kosetsu; Sakai, Tetsuya; Dou, Zhicheng; Tanaka, Katsumi.

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻 8281 LNCS 2013. p. 25-37 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 巻 8281 LNCS).

    研究成果: Conference contribution

    Tsukuda, K, Sakai, T, Dou, Z & Tanaka, K 2013, Estimating intent types for search result diversification. : Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻. 8281 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 巻. 8281 LNCS, pp. 25-37, 9th Asia Information Retrieval Societies Conference on Information Retrieval Technology, AIRS 2013, Singapore, 13/12/9. https://doi.org/10.1007/978-3-642-45068-6_3
    Tsukuda K, Sakai T, Dou Z, Tanaka K. Estimating intent types for search result diversification. : Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻 8281 LNCS. 2013. p. 25-37. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-45068-6_3
    Tsukuda, Kosetsu ; Sakai, Tetsuya ; Dou, Zhicheng ; Tanaka, Katsumi. / Estimating intent types for search result diversification. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻 8281 LNCS 2013. pp. 25-37 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{2a64180520bc412bbc6aba84149ab271,
    title = "Estimating intent types for search result diversification",
    abstract = "Given an ambiguous or underspecified query, search result diversification aims at accommodating different user intents within a single Search Engine Result Page (SERP). While automatic identification of different intents for a given query is a crucial step for result diversification, also important is the estimation of intent types (informational vs. navigational). If it is possible to distinguish between informational and navigational intents, search engines can aim to return one best URL for each navigational intent, while allocating more space to the informational intents within the SERP. In light of the observations, we propose a new framework for search result diversification that is intent importance-aware and type-aware. Our experiments using the NTCIR-9 INTENT Japanese Subtopic Mining and Document Ranking test collections show that: (a) our intent type estimation method for Japanese achieves 64.4{\%} accuracy; and (b) our proposed diversification method achieves 0.6373 in D#-nDCG and 0.5898 in DIN#-nDCG over 56 topics, which are statistically significant gains over the top performers of the NTCIR-9 INTENT Japanese Document Ranking runs. Moreover, our relevance oriented model significantly outperforms our diversity oriented model and the original model by Dou et al..",
    keywords = "Intent Type, Search Result Diversity, Subtopic",
    author = "Kosetsu Tsukuda and Tetsuya Sakai and Zhicheng Dou and Katsumi Tanaka",
    year = "2013",
    doi = "10.1007/978-3-642-45068-6_3",
    language = "English",
    isbn = "9783642450679",
    volume = "8281 LNCS",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    pages = "25--37",
    booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

    }

    TY - GEN

    T1 - Estimating intent types for search result diversification

    AU - Tsukuda, Kosetsu

    AU - Sakai, Tetsuya

    AU - Dou, Zhicheng

    AU - Tanaka, Katsumi

    PY - 2013

    Y1 - 2013

    N2 - Given an ambiguous or underspecified query, search result diversification aims at accommodating different user intents within a single Search Engine Result Page (SERP). While automatic identification of different intents for a given query is a crucial step for result diversification, also important is the estimation of intent types (informational vs. navigational). If it is possible to distinguish between informational and navigational intents, search engines can aim to return one best URL for each navigational intent, while allocating more space to the informational intents within the SERP. In light of the observations, we propose a new framework for search result diversification that is intent importance-aware and type-aware. Our experiments using the NTCIR-9 INTENT Japanese Subtopic Mining and Document Ranking test collections show that: (a) our intent type estimation method for Japanese achieves 64.4% accuracy; and (b) our proposed diversification method achieves 0.6373 in D#-nDCG and 0.5898 in DIN#-nDCG over 56 topics, which are statistically significant gains over the top performers of the NTCIR-9 INTENT Japanese Document Ranking runs. Moreover, our relevance oriented model significantly outperforms our diversity oriented model and the original model by Dou et al..

    AB - Given an ambiguous or underspecified query, search result diversification aims at accommodating different user intents within a single Search Engine Result Page (SERP). While automatic identification of different intents for a given query is a crucial step for result diversification, also important is the estimation of intent types (informational vs. navigational). If it is possible to distinguish between informational and navigational intents, search engines can aim to return one best URL for each navigational intent, while allocating more space to the informational intents within the SERP. In light of the observations, we propose a new framework for search result diversification that is intent importance-aware and type-aware. Our experiments using the NTCIR-9 INTENT Japanese Subtopic Mining and Document Ranking test collections show that: (a) our intent type estimation method for Japanese achieves 64.4% accuracy; and (b) our proposed diversification method achieves 0.6373 in D#-nDCG and 0.5898 in DIN#-nDCG over 56 topics, which are statistically significant gains over the top performers of the NTCIR-9 INTENT Japanese Document Ranking runs. Moreover, our relevance oriented model significantly outperforms our diversity oriented model and the original model by Dou et al..

    KW - Intent Type

    KW - Search Result Diversity

    KW - Subtopic

    UR - http://www.scopus.com/inward/record.url?scp=84893314872&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84893314872&partnerID=8YFLogxK

    U2 - 10.1007/978-3-642-45068-6_3

    DO - 10.1007/978-3-642-45068-6_3

    M3 - Conference contribution

    AN - SCOPUS:84893314872

    SN - 9783642450679

    VL - 8281 LNCS

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    SP - 25

    EP - 37

    BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    ER -