Implicit Preference Labels for Learning Highly Selective Personalized Rankers

Interaction data such as clicks and dwells provide valuable signals for learning and evaluating personalized models. However, while models of personalization typically distinguish between clicked and non-clicked results, no preference distinctions within the non-clicked results are made and all are treated as equally non-relevant. In this paper, we demonstrate that failing to enforce a prior on preferences among non-clicked results leads to learning models that often personalize with no measurable gain at the risk that the personalized ranking is worse than the non-personalized ranking. To address this, we develop an implicit preference-based framework that enables learning highly selective rankers that yield large reductions in risk such as the percentage of queries personalized. We demonstrate theoretically how our framework can be derived from a small number of basic axioms that give rise to well-founded target rankings which combine a weight on prior preferences with the implicit preferences inferred from behavioral data. Additionally, we conduct an empirical analysis to demonstrate that models learned with this approach yield comparable gains on click-based performance measures to standard methods with far fewer queries personalized. On three real-world commercial search engine logs, the method leads to substantial reductions in the number of queries re-ranked (2x - 7x fewer queries re-ranked) while maintaining 85-95% of the total gain achieved by the standard approach.

[1]  Steve Fox,et al.  Evaluating implicit measures to improve web search , 2005, TOIS.

[2]  Qiang Wu,et al.  Adapting boosting for information retrieval measures , 2010, Information Retrieval.

[3]  Ellen M. Voorhees,et al.  TREC 2014 Web Track Overview , 2015, TREC.

[4]  Ellen M. Voorhees,et al.  Overview of the TREC 2014 Web Track , 2017 .

[5]  Susan T. Dumais,et al.  Improving Web Search Ranking by Incorporating User Behavior Information , 2019, SIGIR Forum.

[6]  Thorsten Joachims,et al.  Optimizing search engines using clickthrough data , 2002, KDD.

[7]  Milad Shokouhi,et al.  Fighting search engine amnesia: reranking repeated results , 2013, SIGIR.

[8]  Gleb Gusev,et al.  An Optimization Framework for Weighting Implicit Relevance Labels for Personalized Web Search , 2015, WWW.

[9]  Jaime Teevan,et al.  Understanding and predicting personal navigation , 2011, WSDM '11.

[10]  Olivier Chapelle,et al.  A dynamic bayesian network click model for web search ranking , 2009, WWW '09.

[11]  Qiang Wu,et al.  Learning to Rank Using an Ensemble of Lambda-Gradient Models , 2010, Yahoo! Learning to Rank Challenge.

[12]  Filip Radlinski,et al.  Inferring and using location metadata to personalize web search , 2011, SIGIR.

[13]  Susan T. Dumais,et al.  To personalize or not to personalize: modeling queries with variation in user intent , 2008, SIGIR '08.

[14]  Filip Radlinski,et al.  How does clickthrough data reflect retrieval quality? , 2008, CIKM '08.

[15]  Thorsten Joachims,et al.  Online Structured Prediction via Coactive Learning , 2012, ICML.

[16]  Jaana Kekäläinen,et al.  Cumulated gain-based evaluation of IR techniques , 2002, TOIS.

[17]  Sreenivas Gollapudi,et al.  An axiomatic approach for result diversification , 2009, WWW '09.

[18]  Tie-Yan Liu,et al.  Learning to rank for information retrieval , 2009, SIGIR.

[19]  Kevyn Collins-Thompson,et al.  Reducing the risk of query expansion via robust constrained optimization , 2009, CIKM.

[20]  Milad Shokouhi,et al.  Learning to personalize query auto-completion , 2013, SIGIR.

[21]  Craig MacDonald,et al.  Hypothesis testing for the risk-sensitive evaluation of retrieval systems , 2014, SIGIR.

[22]  Pavel Serdyukov,et al.  Personalization of web-search using short-term browsing context , 2013, CIKM.

[23]  Enhong Chen,et al.  Context-aware ranking in web search , 2010, SIGIR '10.

[24]  M. de Rijke,et al.  Click model-based information retrieval metrics , 2013, SIGIR.

[25]  Fernando Diaz,et al.  Time is of the essence: improving recency ranking using Twitter data , 2010, WWW '10.

[26]  Quoc V. Le,et al.  Learning to Rank with Non-Smooth Cost Functions , 2007 .

[27]  Nick Craswell,et al.  An experimental comparison of click position-bias models , 2008, WSDM '08.

[28]  Wei Chu,et al.  Modeling the impact of short- and long-term behavior on search personalization , 2012, SIGIR '12.

[29]  Filip Radlinski,et al.  Query chains: learning to rank from implicit feedback , 2005, KDD '05.

[30]  Filip Radlinski,et al.  Evaluating the accuracy of implicit feedback from clicks and query reformulations in Web search , 2007, TOIS.

[31]  ChengXiang Zhai,et al.  Semantic term matching in axiomatic approaches to information retrieval , 2006, SIGIR.

[32]  Xuehua Shen,et al.  Context-sensitive information retrieval using implicit feedback , 2005, SIGIR '05.

[33]  Thorsten Joachims,et al.  Stable Coactive Learning via Perturbation , 2013, ICML.

[34]  Susan T. Dumais,et al.  Learning user interaction models for predicting web search result preferences , 2006, SIGIR.

[35]  Benjamin Piwowarski,et al.  A user browsing model to predict search engine click data from past observations. , 2008, SIGIR '08.

[36]  ChengXiang Zhai,et al.  An exploration of axiomatic approaches to information retrieval , 2005, SIGIR '05.

[37]  Bamshad Mobasher,et al.  Web search personalization with ontological user profiles , 2007, CIKM '07.

[38]  Chao Liu,et al.  Click chain model in web search , 2009, WWW '09.

[39]  Maksims Volkovs,et al.  Learning to rank with multiple objective functions , 2011, WWW.

[40]  Paul N. Bennett,et al.  Robust ranking models via risk-sensitive optimization , 2012, SIGIR '12.