Vienna, Austria, 19th September 2015

2nd Workshop on Recommendation Systems for TELEVISION and ONLINE VIDEO

in conjunction with the 9th ACM Conference on Recommender Systems (16th-20th September 2015)

For many households the television is still the central entertainment hub in their home, and the average TV viewer spends about half of their leisure time in front of a TV (3-5 hours/day). The choice of what to watch becomes more overwhelming though because the entertainment options are scattered across various channels, such as on-demand video, digital recorders (on premise or in the cloud) and the traditional linear TV. In addition, consumers can also access the content not just on the big screen, but also on their computers, phones, and tablet devices.
Recommendation systems provide TV users with suggestions about both online video-on-demand and broadcast content and help them to search and browse intelligently for content that is relevant to them.While many open questions in video-on-demand recommendations have already been solved, recommendation systems for broadcast content (e.g., linear channels and catch-up TV) still experience a number of unique challenges due to the peculiarity of such domain. For example, the content available on linear channels is constantly changing and often only available once which leads to severe cold start problems and we often consume TV in groups of varying compositions (household vs individual) which makes building taste profiles and modeling consumer behavior very challenging.
Finally, recommendation systems have to address a number of very different consumption patterns, such as actively browsing through a list of personalized Video on Demand choices that match our current mood, compared to enjoying a “lean back experience” where a recommendation systems playlists a stream of TV shows from our favorite channels for us.
We believe that the combination of the impact of TV and online video focused recommendations on our daily lives together with its challenging nature makes this a very suitable workshop topic for RecSys and of interest for both academic and industrial researchers.

Call for contribution

We would encourage participation along several themes which include but are not limited to:

  • Context-aware TV and online video recommendations
    • Leveraging contextual viewing behaviour, e.g. device specific recommendations
    • Mood based recommendations
    • Group recommendations
  • User modeling & leveraging user viewing and interaction behavior
    • How can social media improve TV recommendations
    • Cross-domain recommendation algorithms (linear TV, video on demand, DVR, gaming consoles)
    • Multi-viewer profile separation
    • Evaluation metrics for TV and online video recommendations
  • Content-based TV and online video recommendations
    • Analysis techniques for video recommendations based on video, audio, or closed caption signals
    • Utilization of external data sources (movie reviews, ratings, plot summaries) for recommendations
  • Other topics related to TV and online video recommendations
    • Video playlisting
    • Linear TV usage and box office success prediction
    • Catch-up TV recommendations
    • Personalized advertisement recommendations
    • Recommendations of 2nd screen web content
    • Recommendations of short form videos (previews, trailers, music videos)

We will welcome works that exploit consumption data of real TV users, with a particular consideration for those releasing the used data set to grant the reproducibility of results and the usage by other researchers.


Jan Neumann, Comcast Labs, Washington, DC (

John Hannon, Zalando SE (

Roberto Turrin, ContentWise, Milan, Italy (

Danny Bickson, Dato, Seattle, WA (

Hassan Sayyadi, Comcast Labs, Washington, DC (

RecSysTV 2015 Schedule

9:00 – 9:10: Opening Introduction

9:10 – 09:50: Invited Talk 1 – Recommending TV News and Circumventing The Filter Bubble – Daniel Salas (Thomson Reuters)

Reuters TV is an on-demand service that offers personalized news coverage straight from the source. To live up to this offering, we not only have the challenge of learning the preferences of each individual user but also of abiding to editorial constraints to avoid the creation of a filter bubble.
In this talk, we outline our recommendation framework and discuss our approach to modeling user preferences, engineering feature representations for video news stories, combining collaborative and content-based recommendation, and applying approximate optimal control to news program assembly under editorial ranking constraints.

9:50 – 10:30: Invited Talk 2 – From the ground up: A journey on building a recommender system for a video mobile app before a product launch – Diana Hu & Joaquin Delgado, OnCue TV (Verizon)

Recommendation engines have a proven track record of success when a similarity index between users and items can be built. This depends upon the presence of a stable item catalog and the availability of a feedback signal with historical data on the users’ preferences; however, when these signals are not available, conventional solutions, such as collaborative filtering, face the cold start problem. In addition, the item catalog for our video mobile app contains non-traditional content that is ever changing. This brings a new set of challenges and complexities. On this talk, we will present our journey on how we solved this problem with a combination of strong engineering, practical data science, and balanced product design.

10:30 – 11:00: Morning Coffee Break

11:00 – 11:55: Session 1 – Social Sources for Television Recommenders

  • Relevance of Social Data in Video Recommendation (25 min)pdf
  • Using Social Media data for Online Television Recommendation Services at RTÉ Ireland (15 min)pdf
  • Exploiting crowdsourced movie reviews to explain recommendation (15 min)pdf

11:55 – 12:30: Company Demos: Dato, ContentWise, Demo Lighting Round

12:30 – 14:00: Lunch with Poster and Demo Session over lunch break

14:00 – 14:50: Invited Talk 3 – How smart is your data? The new differentiator for video operators – Thibault D’Orso (

14:50 – 15:30: Session 2 – Evaluation Techniques for Television Recommenders

  • Context-aware LDA: Balancing Relevance and Diversity in TV Content Recommenders (25 min)pdf
  • New Quality Measure of Linear Ads in Online Videos (15 min)pdf

15:30 – 16:00: Afternoon Coffee Break

16:00 – 16:50: Session 3 – Mathematical models for TV and Short Video Recommendations

  • Prediction of TV ratings with dynamic models (25 min)pdf
  • An Adaptive Implicit Feedback Model for Short Clips Recommendations (25 min)pdf

16:50 – 17:00: Wrap up

Important dates

Submission deadline: July 5, 2015

Notification: July 20, 2015

Camera-ready: August 4, 2015

Deadline for author registration: August 16, 2015

Workshop date: September 19, 2015 (full day)


We are soliciting submissions of long and short papers, as well as position presentations.
Long paper are to represent original mature research and can be 6-8 pages long. We request potential submitters to adhere to double-column ACM SIG format in line with standard RecSys formatting guidelines.
Short papers are to represent early/promising research, demos or industrial case studies and can be 4 pages in length (ACM RecSys style) or up to 20 slides.
Use the following website to electronically submit your paper:
Note that attendance at the workshop requires registration for the ACM RecSys 2015 conference as a whole. This year there is no separate registration for workshops. Each accepted workshop paper must register at least one author at the conference.

Program commitee

Hidasi Balazs, GravityR&D

Justin Basilico, Netflix

Craig Carmichael, Rovi

Emanuele Coviello, Keevio

Paolo Cremonesi, Politecnico di Milano

Joaquin Delgado, OnCue TV (Verizon)

Diana Hu, OnCue TV (Verizon)

Brendan Kitts, Adapt.TV (AOL)

Gert Lanckriet, UC San Diego

Royi Ronen, Microsoft

Barry Smyth, Insight Centre for Data Analytics

Esti Widder, Viaccess-Orca

Jiayu Zhou, Samsung Research

David Zibriczky, ImpressTv