Optimizing the Recency-Relevancy Trade-off in Online News Recommendations

Online news media sites are emerging as the primary source of news for a large number of users. The selection of 'front-page' stories on these media sites usually takes into consideration several crowdsourced popularity metrics, such as number of views or shares by the readers. In this work, we focus on automatically recommending front-page stories in such media websites. When recommending news stories, there are two basic metrics of interest - recency and relevancy. Ideally, recommender systems should recommend the most relevant stories soon after they are published. However, the relevancy of a story only becomes evident as the story ages, thereby creating a tension between recency and relevancy. A systematic analysis of popular recommendation strategies in use today reveals that they lead to poor trade-offs between recency and relevancy in practice. So, in this paper, we propose a new recommendation strategy (called Highest Future-Impact) which attempts to optimize on both the axes. To implement our proposed strategy in practice, we develop an optimization framework combining the predicted future-impact of the stories with the uncertainties in the predictions. Evaluations over three real-world news datasets show that our implementation achieves good performance trade-offs between recency and relevancy.

[1]  Jiawei Han,et al.  Citation Prediction in Heterogeneous Bibliographic Networks , 2012, SDM.

[2]  R. Koenker.  Quantile Regression: Name Index , 2005 .

[3]  Kazufumi Watanabe,et al.  Jasmine: a real-time local-event detection system based on geolocation information propagated to microblogs , 2011, CIKM '11.

[4]  Julian J. Faraway,et al.  Practical Regression and Anova using R , 2002 .

[5]  Sean J. Taylor,et al.  Social Influence Bias: A Randomized Experiment , 2013, Science.

[6]  Leo Breiman,et al.  Bias, Variance , And Arcing Classifiers , 1996 .

[7]  Balaji Padmanabhan,et al.  SCENE: a scalable two-stage personalized news recommendation system , 2011, SIGIR.

[8]  Naomi S. Altman,et al.  Quantile regression , 2019, Nature Methods.

[9]  E. Katz.  The Two-Step Flow of Communication: An Up-To-Date Report on an Hypothesis , 1957 .

[10]  Jürgen Habermas,et al.  The Public Sphere , 2018, Princeton Readings in Political Thought.

[11]  Joemon M. Jose,et al.  "Nobody comes here anymore, it's too crowded"; Predicting Image Popularity on Flickr , 2014, ICMR.

[12]  Jürgen Habermas,et al.  The Public Sphere: An Encyclopedia Article (1964) , 1974 .

[13]  Nick Koudas,et al.  TwitterMonitor: trend detection over the twitter stream , 2010, SIGMOD Conference.

[14]  Marsha L. Richins,et al.  The Role of Evolvement and Opinion Leadership in Consumer Word-Of-Mouth: an Implicit Model Made Explicit , 1988 .

[15]  Krishna P. Gummadi,et al.  Dissemination Biases of Social Media Channels: On the Topical Coverage of Socially Shared News , 2016, ICWSM.

[16]  Yue Xu,et al.  Time-aware topic recommendation based on micro-blogs , 2012, CIKM.

[17]  Deepak Agarwal,et al.  Click shaping to optimize multiple objectives , 2011, KDD.

[18]  Júlio Cesar dos Reis,et al.  Breaking the News: First Impressions Matter on Online News , 2015, ICWSM.

[19]  Bernardo A. Huberman,et al.  The Pulse of News in Social Media: Forecasting Popularity , 2012, ICWSM.

[20]  Flavio Figueiredo,et al.  On the Dynamics of Social Media Popularity: A YouTube Case Study , 2014, TOIT.

[21]  Wei Chu,et al.  Unbiased offline evaluation of contextual-bandit-based news article recommendation algorithms , 2010, WSDM '11.

[22]  References , 1971 .

[23]  Jie Tang,et al.  Citation count prediction: learning to estimate future citations for literature , 2011, CIKM '11.

[24]  Boi Faltings,et al.  Predicting Online Performance of News Recommender Systems Through Richer Evaluation Metrics , 2015, RecSys.

[25]  Krishna P. Gummadi,et al.  Can Trending News Stories Create Coverage Bias? On the Impact of High Content Churn in Online News Media , 2015 .

[26]  Jiahui Liu,et al.  Personalized news recommendation based on click behavior , 2010, IUI '10.

[27]  Niloy Ganguly,et al.  Stop Clickbait: Detecting and preventing clickbaits in online news media , 2016, 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM).

[28]  Martha Larson,et al.  Benchmarking News Recommendations: The CLEF NewsREEL Use Case , 2016, SIGF.

[29]  P. J. Tichenor,et al.  MASS MEDIA FLOW AND DIFFERENTIAL GROWTH IN KNOWLEDGE , 1970 .

[30]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[31]  Didier Sornette,et al.  Robust dynamic classes revealed by measuring the response function of a social system , 2008, Proceedings of the National Academy of Sciences.

[32]  Wei-Yin Loh,et al.  Classification and regression trees , 2011, WIREs Data Mining Knowl. Discov..

[33]  C. Dunnett,et al.  A BIVARIATE GENERALIZATION OF STUDENT'S t-DISTRIBUTION, WITH TABLES FOR CERTAIN SPECIAL CASES , 1954 .

[34]  Ciro Cattuto,et al.  Dynamical classes of collective attention in twitter , 2011, WWW.

[35]  Matthew J. Salganik,et al.  Leading the Herd Astray: An Experimental Study of Self-fulfilling Prophecies in an Artificial Cultural Market , 2008, Social psychology quarterly.