Analyzing the Interaction of Users with News Articles to Create Personalization Services

News personalization technologies aim at providing contents tailored to the users preferences. While recommender systems provide suggestions to users by taking advantage of individual preferences, content of news portals can be tailored on the bases of sociological aspects (e.g., the demographics of the users or the region in which they live) elicited from user interactions with the news. This allows to generate personalization with a coarse granularity; however, no study has ever shown a large-scale analysis of how the users interact with the news, focusing on different user segments. This paper uses the Yahoo News Feed dataset, a corpus that contains more than 101 billion examples of interactions between users and news items. The data present in the corpus spans for a range of 4 months, and was extracted from real user interactions with news items in the Yahoo Web portal. The dataset has been analyzed in order to understand users behaviors and their relations with sociological aspects. Thanks to our analysis, different forms of personalization can be generated.

[1]  Ludovico Boratto,et al.  Modeling the Preferences of a Group of Users Detected by Clustering: a Group Recommendation Case-Study , 2014, WIMS '14.

[2]  M. Angela Sasse,et al.  Too close for comfort: a study of the effectiveness and acceptability of rich-media personalized advertising , 2012, CHI.

[3]  Taghi M. Khoshgoftaar,et al.  A Survey of Collaborative Filtering Techniques , 2009, Adv. Artif. Intell..

[4]  Ludovico Boratto,et al.  The rating prediction task in a group recommender system that automatically detects groups: architectures, algorithms, and performance evaluation , 2014, Journal of Intelligent Information Systems.

[5]  Uzay Kaymak,et al.  News personalization using the CF-IDF semantic recommender , 2011, WIMS '11.

[6]  A. Barabasi,et al.  Dynamics of information access on the web. , 2006, Physical review. E, Statistical, nonlinear, and soft matter physics.

[7]  Pasquale Lops,et al.  Content-based Recommender Systems: State of the Art and Trends , 2011, Recommender Systems Handbook.

[8]  Shunzhi Zhu,et al.  Personalized News Recommendation: A Review and an Experimental Investigation , 2011, Journal of Computer Science and Technology.

[9]  Ludovico Boratto,et al.  ART: group recommendation approaches for automatically detected groups , 2015, Int. J. Mach. Learn. Cybern..

[10]  Karl Pearson F.R.S..  X. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling , 2009 .

[11]  Lawrence Birnbaum,et al.  Social media-driven news personalization , 2012, RSWeb@RecSys.

[12]  K. Pearson.  On the Criterion that a Given System of Deviations from the Probable in the Case of a Correlated System of Variables is Such that it Can be Reasonably Supposed to have Arisen from Random Sampling , 1900 .

[13]  Paola Velardi,et al.  What to Write? A topic recommender for journalists , 2017, NLPmJ@EMNLP.

[14]  Veda C. Storey,et al.  Business Intelligence and Analytics: From Big Data to Big Impact , 2012, MIS Q..

[15]  Ludovico Boratto,et al.  A Tool to Analyze the Reading Behavior of the Users in a Mobile Digital Publishing Platform , 2016, KDWeb.

[16]  Deepak Agarwal,et al.  Multi-faceted ranking of news articles using post-read actions , 2012, CIKM '12.

[17]  Abhinandan Das,et al.  Google news personalization: scalable online collaborative filtering , 2007, WWW '07.

[18]  Jürgen Pfeffer,et al.  Characterizing the life cycle of online news stories using social media reactions , 2013, CSCW.

[19]  Francesco Gullo,et al.  Exploiting search history of users for news personalization , 2017, Inf. Sci..