ABSTRACT
Easy accessibility can often lead to over-consumption, as seen in food and alcohol habits. On video on-demand (VOD) services, this has recently been referred to as binge watching, where potentially entire seasons of TV shows are consumed in a single viewing session. While a user viewership model may reveal this binging behavior, creating an accurate model has several challenges, including censored data, deviations in the population, and the need to consider external influences on consumption habits. In this paper, we introduce a novel statistical mixture model that incorporates these factors and presents a first of its kind characterization of viewer consumption behavior using a real-world dataset that includes playback data from a VOD service. From our modeling, we tackle various predictive tasks to infer the consumption decisions of a user in a viewing session, including estimating the number of episodes they watch and classifying if they continue watching another episode. Using these insights, we then identify binge watching sessions based on deviation from normal viewing behavior. We observe different types of binging behavior, that binge watchers often view certain content out-of-order, and that binge watching is not a consistent behavior among our users. These insights and our findings have application in VOD revenue generation, consumer health applications, and customer retention analysis.
Supplemental Material
- Alcohol Research Group Datasets. http://arg.org/resources-tools/databases/.Google Scholar
- The Transparency Project. http://www.thetransparencyproject.org/.Google Scholar
- Netflix declares binge watching is the new normal. http://www.prnewswire.com/news-releases/netflix-declares-binge-watching-is-the-new-normal-235713431.html, 2013.Google Scholar
- H. Abrahamsson and M. Nordmark. Program popularity and viewer behavior in a large TV-on-demand system. In ACM Internet Measurement Conference, pages 199--210, 2012. Google ScholarDigital Library
- C. M. Bishop. Pattern Recognition and Machine Learning. Springer-Verlag New York, 2006. Google ScholarDigital Library
- L. Carlinet, T. Huynh, B. Kauffmann, F. Mathieu, L. Noirie, and S. Tixeuil. Four months in daily motion: Dissecting user video requests. In IEEE Wireless Communications and Mobile Computing Conference, pages 613--618, 2012.Google ScholarCross Ref
- M. Cha, H. Kwak, P. Rodriguez, Y.-Y. Ahn, and S. Moon. Analyzing the video popularity characteristics of large-scale user generated content systems. IEEE/ACM Transactions on Networking, 17(5):1357--1370, 2009. Google ScholarDigital Library
- A. Dempster, N. Laird, and D. Rubin. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, pages 1--38, 1977.Google ScholarCross Ref
- G. Dupret and M. Lalmas. Absence time and user engagement: evaluating ranking functions. In ACM international conference on Web search and Data Mining, pages 173--182, 2013. Google ScholarDigital Library
- L. A. Hannah, D. M. Blei, and W. B. Powell. Dirichlet process mixtures of generalized linear models. The Journal of Machine Learning Research, 12:1923--1953, 2011. Google ScholarDigital Library
- Y. He and K. Wang. Inferring search behaviors using partially observable markov model with duration. In Proceedings of the fourth ACM International Conference on Web search and Data Mining, pages 415--424, 2011. Google ScholarDigital Library
- K. Kapoor, M. Sun, J. Srivastava, and T. Ye. A hazard based approach to user return time prediction. In ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 1719--1728, 2014. Google ScholarDigital Library
- D. Karlis, P. Papatla, and S. Roy. Finite mixtures of censored Poisson regression models. Statistica Neerlandica, 2015.Google Scholar
- R. Kay. Proportional hazard regression models and the analysis of censored survival data. Applied Statistics, pages 227--237, 1977.Google Scholar
- C. McCulloch and S. Searle. Generalized linear and mixed Models. John Wiley and Sons, 2004.Google Scholar
- S. Nagano, Y. Ichikawa, N. Takaya, T. Uchiyama, and M. Abe. Nonparametric hierarchal Bayesian modeling in non-contractual heterogeneous survival data. In ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 668--676, 2013. Google ScholarDigital Library
- TiVO. Viewers' New TV Reality: A "Season" = One Day. http://pr.tivo.com/press-releases, 2014.Google Scholar
- R. A. Volberg. The prevalence and demographics of pathological gamblers: implications for public health. American Journal of Public Health, 84(2):237--241, 1994.Google ScholarCross Ref
- G. Willmot. Mixed compound Poisson distributions. Astin Bulletin, 16(S1):S59--S79, 1986.Google Scholar
- W. C.-H. Wu, M.-Y. Yeh, and M.-S. Chen. Predicting winning price in real time bidding with censored data. In ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 1305--1314, 2015. Google ScholarDigital Library
- H. Yeh, B. Gajewski, P. Mukhopadhyay, and F. Behbod. The zero-truncated Poisson with right censoring: an application to translational breast cancer research. Statistics in Biopharmaceutical Research, 4(3):252--263, 2012.Google ScholarCross Ref
- H. Yu, D. Zheng, B. Y. Zhao, and W. Zheng. Understanding user behavior in large-scale video-on-demand systems. In ACM SIGOPS Operating Systems Review, volume 40, pages 333--344, 2006. Google ScholarDigital Library
Index Terms
- Just One More: Modeling Binge Watching Behavior
Recommendations
"I Can Watch What I Want": A Diary Study of On-Demand and Cross-Device Viewing
TVX '18: Proceedings of the 2018 ACM International Conference on Interactive Experiences for TV and Online VideoIn recent years, on-demand video services, such as Netflix and Amazon Video, have become extremely popular. To understand how people use these services, we recruited 20 people from nine households to keep a viewing diary for 14 days. To better ...
Netflix and Ill?: Emotional and Health Implications of Binge Watching Streaming TV
SMSociety '18: Proceedings of the 9th International Conference on Social Media and SocietyIt has been suggested in some preliminary and somewhat anecdotal work that the effects of binge watching platforms such as Netflix, Hulu and Amazon Prime have been reported to include depression, chronic illness, weight gain, sleep disorders, and a ...
To Binge or not to Binge: Viewers’ Moods and Behaviors During the Consumption of Subscribed Video Streaming
Entertainment Computing – ICEC 2020AbstractThe popularity of internet-distributed TV entertainment services, such as Netflix, has transformed TV consumption behavior. Currently, the level of control viewers have over their TV experiences, along with the release of complete seasons at once, ...
Comments