Counterfactual Learning To Rank for Utility-Maximizing Query Autocompletion

被引:0
|
作者
Block, Adam [1 ,5 ]
Kidambi, Rahul [2 ]
Hill, Daniel N. [3 ]
Joachims, Thorsten [4 ,5 ]
机构
[1] MIT, 77 Massachusetts Ave, Cambridge, MA 02139 USA
[2] Amazon Search, Berkeley, CA USA
[3] Amazon Mus, San Francisco, CA USA
[4] UT Austin, Austin, TX USA
[5] Amazon, Seattle, WA USA
基金
美国国家科学基金会;
关键词
Query Auto-Complete; Learning to Rank; Counterfactual Estimation;
D O I
10.1145/3477495.3531958
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Conventional methods for query autocompletion aim to predict which completed query a user will select from a list. A shortcoming of this approach is that users often do not know which query will provide the best retrieval performance on the current information retrieval system, meaning that any query autocompletion methods trained to mimic user behavior can lead to suboptimal query suggestions. To overcome this limitation, we propose a new approach that explicitly optimizes the query suggestions for downstream retrieval performance. We formulate this as a problem of ranking a set of rankings, where each query suggestion is represented by the downstream item ranking it produces. We then present a learning method that ranks query suggestions by the quality of their item rankings. The algorithm is based on a counterfactual learning approach that is able to leverage feedback on the items (e.g., clicks, purchases) to evaluate query suggestions through an unbiased estimator, thus avoiding the assumption that users write or select optimal queries. We establish theoretical support for the proposed approach and provide learning-theoretic guarantees. We also present empirical results on publicly available datasets, and demonstrate real-world applicability using data from an online shopping store.
引用
收藏
页码:791 / 802
页数:12
相关论文
共 50 条
  • [1] Causal Feature Learning for Utility-Maximizing Agents
    Kinney, David
    Watson, David
    [J]. INTERNATIONAL CONFERENCE ON PROBABILISTIC GRAPHICAL MODELS, VOL 138, 2020, 138 : 257 - 268
  • [2] Utility-maximizing Server Selection
    Truong Khoa Phan
    Griffin, David
    Maini, Elisa
    Rio, Miguel
    [J]. 2016 IFIP NETWORKING CONFERENCE (IFIP NETWORKING) AND WORKSHOPS, 2016, : 413 - 421
  • [3] UNIVERSALLY UTILITY-MAXIMIZING PRIVACY MECHANISMS
    Ghosh, Arpita
    Roughgarden, Tim
    Sundararajan, Mukund
    [J]. SIAM JOURNAL ON COMPUTING, 2012, 41 (06) : 1673 - 1693
  • [4] STAGGERED PRICES IN A UTILITY-MAXIMIZING FRAMEWORK
    CALVO, GA
    [J]. JOURNAL OF MONETARY ECONOMICS, 1983, 12 (03) : 383 - 398
  • [5] Universally Utility-Maximizing Privacy Mechanisms
    Ghosh, Arpita
    Roughgarden, Tim
    Sundararajan, Mukund
    [J]. STOC'09: PROCEEDINGS OF THE 2009 ACM SYMPOSIUM ON THEORY OF COMPUTING, 2009, : 351 - 359
  • [6] OPTIMAL-CONTRACTS WITH A UTILITY-MAXIMIZING AUDITOR
    BAIMAN, S
    EVANS, JH
    NOEL, J
    [J]. JOURNAL OF ACCOUNTING RESEARCH, 1987, 25 (02) : 217 - 244
  • [7] A UTILITY-MAXIMIZING MECHANISM FOR VICARIOUS REWARD - COMMENTS
    AINSLIE, G
    [J]. RATIONALITY AND SOCIETY, 1995, 7 (04) : 393 - 403
  • [8] Model selection in utility-maximizing binary prediction
    Su, Jiun-Hua
    [J]. JOURNAL OF ECONOMETRICS, 2021, 223 (01) : 96 - 124
  • [9] The utility-maximizing self-employed physician
    Thornton, J
    Eakin, BK
    [J]. JOURNAL OF HUMAN RESOURCES, 1997, 32 (01) : 98 - 128
  • [10] Robust Generalization and Safe Query-Specialization in Counterfactual Learning to Rank
    Oosterhuis, Harrie
    de Rijke, Maarten
    [J]. PROCEEDINGS OF THE WORLD WIDE WEB CONFERENCE 2021 (WWW 2021), 2021, : 158 - 170