Improving Post-Click User Engagement on Native Ads via Survival Analysis

In this paper we focus on estimating the post-click engagement on native ads by predicting the dwell time on the corresponding ad landing pages. To infer relationships between features of the ads and dwell time we resort to the application of survival analysis techniques, which allow us to estimate the distribution of the length of time that the user will spend on the ad. This information is then integrated into the ad ranking function with the goal of promoting the rank of ads that are likely to be clicked and consumed by users (dwell time greater than a given threshold). The online evaluation over live traffic shows that considering post-click engagement has a consistent positive effect on both CTR, decreases the number of bounces and increases the average dwell time, hence leading to a better user post-click experience.

[1]  Xiaoli Z. Fern,et al.  Visual appearance of display ads and its effect on click through rate , 2012, CIKM.

[2]  Wayne Nelson.  Theory and Applications of Hazard Plotting for Censored Failure Data , 2000, Technometrics.

[3]  SangKeun Lee,et al.  Semantic contextual advertising based on the open directory project , 2013, TWEB.

[4]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[5]  Diane Tang,et al.  Focusing on the Long-term: It's Good for Users and Business , 2015, KDD.

[6]  W. Chu.  Studying Aesthetics in Photographic Images Using a Computational Approach , 2013 .

[7]  Haibin Liu,et al.  Characterizing Landing Pages in Sponsored Search , 2012, 2012 Eighth Latin American Web Congress.

[8]  Suju Rajan,et al.  Beyond clicks: dwell time for personalization , 2014, RecSys '14.

[9]  Fabrizio Silvestri,et al.  Promoting Positive Post-Click Experience for In-Stream Yahoo Gemini Users , 2015, KDD.

[10]  D. Sculley,et al.  Predicting bounce rates in sponsored search advertisements , 2009, KDD.

[11]  David Lo,et al.  Predicting response in mobile advertising with hierarchical importance-aware factorization machine , 2014, WSDM.

[12]  E. Kaplan,et al.  Nonparametric Estimation from Incomplete Observations , 1958 .

[13]  D. Kleinbaum,et al.  Survival Analysis: A Self-Learning Text. , 1996 .

[14]  Elizabeth F. Churchill,et al.  Mobile advertising: evaluating the effects of animation, user and content relevance , 2013, CHI.

[15]  Rómer Rosales,et al.  Post-click conversion modeling and analysis for non-guaranteed delivery display advertising , 2012, WSDM '12.

[16]  Dragomir Yankov,et al.  Categorization of display ads using image and landing page features , 2011, LDMTA '11.

[17]  Hila Becker,et al.  Context transfer in search advertising , 2009, SIGIR.

[18]  Hemant Ishwaran,et al.  Random Survival Forests , 2008, Wiley StatsRef: Statistics Reference Online.

[19]  Kathryn E. Boudreau.  Mobile Advertising and its Acceptance by American Consumers , 2013 .

[20]  M. Pagano,et al.  Survival analysis. , 1996, Nutrition.

[21]  Ron Kohavi,et al.  Trustworthy online controlled experiments: five puzzling outcomes explained , 2012, KDD.

[22]  Peifeng Yin,et al.  Silence is also evidence: interpreting dwell time for recommendation from psychological perspective , 2013, KDD.

[23]  Andrei Z. Broder,et al.  A semantic approach to contextual advertising , 2007, SIGIR.

[24]  Raymond K. Wong,et al.  Framework for timely and accurate ads on mobile devices , 2009, CIKM.

[25]  Hila Becker,et al.  What happens after an ad click?: quantifying the impact of landing pages in web advertising , 2009, CIKM.

[26]  Dustin Hillard,et al.  A predictive model for advertiser value-per-click in sponsored search , 2013, WWW.

[27]  Yejin Choi,et al.  Using landing pages for sponsored search ad selection , 2010, WWW '10.

[28]  Mounia Lalmas,et al.  Absence time and user engagement: evaluating ranking functions , 2013, WSDM '13.

[29]  D.,et al.  Regression Models and Life-Tables , 2022 .

[30]  Vassilis Plachouras,et al.  A noisy-channel approach to contextual advertising , 2007, ADKDD '07.

[31]  Ryen W. White,et al.  Modeling dwell time to predict click-level satisfaction , 2014, WSDM.