Workload-driven learning of mallows mixtures with pairwise preference data

Julia Stoyanovich, Lovro Ilijasic, Haoyue Ping

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

    Abstract

    In this paper we present a framework for learning mixtures of Mallows models from large samples of incomplete preferences. The problem we address is of significant practical importance in social choice, recommender systems, and other domains where it is required to aggregate, or otherwise analyze, preferences of a heterogeneous user base. We improve on state-of-the-art methods for learning mixtures of Mallows models with pairwise preference data. Exact sampling from the Mallows posterior in presence of arbitrary pairwise evidence is known to be intractable even for a single Mallows. This motivated to the development of an approximate sampler called AMP. In this paper we propose AMPx, an ensemble method for approximate sampling from the Mallows posterior that combines AMP with frequency-based estimation of posterior probabilities. We experimentally demonstrate that AMPx achieves faster convergence and higher accuracy than AMP alone. We also adapt stateof-the-art clustering techniques that have not been used in this setting, for learning parameters of the Mallows mixture, and show experimentally that mixture parameters can be learned accurately and efficiently.

    Original languageEnglish (US)
    Title of host publicationProceedings of the 19th International Workshop on Web and Databases, WebDB 2016
    PublisherAssociation for Computing Machinery, Inc
    ISBN (Electronic)9781450343107
    DOIs
    StatePublished - Jun 26 2016
    Event19th International Workshop on Web and Databases, WebDB 2016 - San Francisco, United States
    Duration: Jun 26 2016Jul 1 2016

    Other

    Other19th International Workshop on Web and Databases, WebDB 2016
    CountryUnited States
    CitySan Francisco
    Period6/26/167/1/16

    Fingerprint

    Sampling
    Recommender systems

    ASJC Scopus subject areas

    • Computer Networks and Communications
    • Computer Science Applications
    • Information Systems

    Cite this

    Stoyanovich, J., Ilijasic, L., & Ping, H. (2016). Workload-driven learning of mallows mixtures with pairwise preference data. In Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016 Association for Computing Machinery, Inc. https://doi.org/10.1145/2932194.2932202

    Workload-driven learning of mallows mixtures with pairwise preference data. / Stoyanovich, Julia; Ilijasic, Lovro; Ping, Haoyue.

    Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016. Association for Computing Machinery, Inc, 2016.

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

    Stoyanovich, J, Ilijasic, L & Ping, H 2016, Workload-driven learning of mallows mixtures with pairwise preference data. in Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016. Association for Computing Machinery, Inc, 19th International Workshop on Web and Databases, WebDB 2016, San Francisco, United States, 6/26/16. https://doi.org/10.1145/2932194.2932202
    Stoyanovich J, Ilijasic L, Ping H. Workload-driven learning of mallows mixtures with pairwise preference data. In Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016. Association for Computing Machinery, Inc. 2016 https://doi.org/10.1145/2932194.2932202
    Stoyanovich, Julia ; Ilijasic, Lovro ; Ping, Haoyue. / Workload-driven learning of mallows mixtures with pairwise preference data. Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016. Association for Computing Machinery, Inc, 2016.
    @inproceedings{ed37e365fee5486ba25a63a0fd96a264,
    title = "Workload-driven learning of mallows mixtures with pairwise preference data",
    abstract = "In this paper we present a framework for learning mixtures of Mallows models from large samples of incomplete preferences. The problem we address is of significant practical importance in social choice, recommender systems, and other domains where it is required to aggregate, or otherwise analyze, preferences of a heterogeneous user base. We improve on state-of-the-art methods for learning mixtures of Mallows models with pairwise preference data. Exact sampling from the Mallows posterior in presence of arbitrary pairwise evidence is known to be intractable even for a single Mallows. This motivated to the development of an approximate sampler called AMP. In this paper we propose AMPx, an ensemble method for approximate sampling from the Mallows posterior that combines AMP with frequency-based estimation of posterior probabilities. We experimentally demonstrate that AMPx achieves faster convergence and higher accuracy than AMP alone. We also adapt stateof-the-art clustering techniques that have not been used in this setting, for learning parameters of the Mallows mixture, and show experimentally that mixture parameters can be learned accurately and efficiently.",
    author = "Julia Stoyanovich and Lovro Ilijasic and Haoyue Ping",
    year = "2016",
    month = "6",
    day = "26",
    doi = "10.1145/2932194.2932202",
    language = "English (US)",
    booktitle = "Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016",
    publisher = "Association for Computing Machinery, Inc",

    }

    TY - GEN

    T1 - Workload-driven learning of mallows mixtures with pairwise preference data

    AU - Stoyanovich, Julia

    AU - Ilijasic, Lovro

    AU - Ping, Haoyue

    PY - 2016/6/26

    Y1 - 2016/6/26

    N2 - In this paper we present a framework for learning mixtures of Mallows models from large samples of incomplete preferences. The problem we address is of significant practical importance in social choice, recommender systems, and other domains where it is required to aggregate, or otherwise analyze, preferences of a heterogeneous user base. We improve on state-of-the-art methods for learning mixtures of Mallows models with pairwise preference data. Exact sampling from the Mallows posterior in presence of arbitrary pairwise evidence is known to be intractable even for a single Mallows. This motivated to the development of an approximate sampler called AMP. In this paper we propose AMPx, an ensemble method for approximate sampling from the Mallows posterior that combines AMP with frequency-based estimation of posterior probabilities. We experimentally demonstrate that AMPx achieves faster convergence and higher accuracy than AMP alone. We also adapt stateof-the-art clustering techniques that have not been used in this setting, for learning parameters of the Mallows mixture, and show experimentally that mixture parameters can be learned accurately and efficiently.

    AB - In this paper we present a framework for learning mixtures of Mallows models from large samples of incomplete preferences. The problem we address is of significant practical importance in social choice, recommender systems, and other domains where it is required to aggregate, or otherwise analyze, preferences of a heterogeneous user base. We improve on state-of-the-art methods for learning mixtures of Mallows models with pairwise preference data. Exact sampling from the Mallows posterior in presence of arbitrary pairwise evidence is known to be intractable even for a single Mallows. This motivated to the development of an approximate sampler called AMP. In this paper we propose AMPx, an ensemble method for approximate sampling from the Mallows posterior that combines AMP with frequency-based estimation of posterior probabilities. We experimentally demonstrate that AMPx achieves faster convergence and higher accuracy than AMP alone. We also adapt stateof-the-art clustering techniques that have not been used in this setting, for learning parameters of the Mallows mixture, and show experimentally that mixture parameters can be learned accurately and efficiently.

    UR - http://www.scopus.com/inward/record.url?scp=84979775163&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84979775163&partnerID=8YFLogxK

    U2 - 10.1145/2932194.2932202

    DO - 10.1145/2932194.2932202

    M3 - Conference contribution

    BT - Proceedings of the 19th International Workshop on Web and Databases, WebDB 2016

    PB - Association for Computing Machinery, Inc

    ER -