Overall winner: Yue Shi, Martha Larson and Alan Hanjalic. Mining Mood-specific Movie Similarity with Matrix Factorization for Context-aware Recommendation

Weekly track winner: Zeno Gantner, Steffen Rendle and Lars Schmidt-Thieme. Time vs. Tags: Factorization Models for Context-/Time-Aware Movie Recommendations

Mood track winner: Yue Shi, Martha Larson and Alan Hanjalic. Mining Mood-specific Movie Similarity with Matrix Factorization for Context-aware Recommendation

Social track winner: Nathan Liu.  Adapting Neighborhood and Matrix Factorization Models for Context Aware Recommendations

Live track winner: Zeno Gantner, Steffen Rendle and Lars Schmidt-Thieme. Time vs. Tags: Factorization Models for Context-/Time-Aware Movie Recommendations

Thanks to all of you participated!

Gedas Adomavicius said anyone interested in joining the list should send him an email at gedas [at] umn [dot] edu

We’re planning to ask these questions:

 
We’ll send out the survey this afternoon and would love to get your feedback before it goes out.

Which questions are we missing? Do we ask the right questions? Please comment!

A map is available here, click “A Rooms” for an overview.

The Context of a Recommendation

Abstract: Recommender Systems are software tools and techniques providing suggestions for items to be of use to a user. Context is providing information that can influence the perception of the usefulness of an item for a user. For this reason Recommender Systems must take into account this information to deliver more useful (perceived) recommendations. There are several examples motivating the importance of context for Recommender Systems. For instance, to suggest a meaningful travel to a user a Recommender System must know if the travel is scheduled in summer or winter, and if the user is traveling alone or with kids. Moreover, only considering the context of an evaluation – the user rating for an item – one can assign a proper meaning to that. For instance while one can judge a Ferrari a 5 star car; this does not mean that it is a proper recommendation if the user is looking for a new car to buy. Context modeling and context-dependent reasoning is a complex subject and still there are major technical and practical difficulties to solve: obtain sufficient and reliable data describing the user preferences in context; selecting the right context information, i.e., relevant in a particular personalization task; understanding the impact of the contextual dimensions on the personalization process; embedding the contextual dimensions in a recommendation computational model. Besides, Recommender Systems research should consider the wider scope of contextual computing, defined as the enhancement of the user’s interactions by understanding the user, the context, and the applications and information being used, typically across a wide set of user goals. Actively adapting the computational environment – for each user – at each point of computation should be the ultimate goal. So, contextual computing for Recommender Systems must also focus on understanding the information consumption patterns of each user, and on the process not only on the output of the recommendation process.

Submit you papers here

We will be posting a CFP when all details have been sorted out.