Section: Application Domains

Recommendation systems in a broad sense

Recommendation systems have been a major field of applications of our research for a few years now. Recommendation systems should be understood in a broad sense, as systems that aim at providing personalized responses/items to users, based on their characteristics, and the environment in which the interaction happens.

In that broad sense, we have collaborated with companies on computational advertizing and recommendation systems. These collaborations have involved research studies on the following issues:

  • cold-start problem,

  • time varying environment,

  • ability to deal with large amounts of users and items,

  • ability to design algorithms to respond within a reasonnable amount of time, usually below 1 millisecond.

We have also competed in challenges, winning some of them (SequeL ranked first and second at the “Pascal Exploration & Exploitation Challenge 2011”; SequeL ranked first at the “RecSys Challenge 2014: User Engagement as Evaluation”.), and we have also organized a challenge (ICML 2012 new Challenges for Exploration & Exploitation 3.), on those topics.

A company has been awarded an innovation award in 2015, thanks to the research work done in collaboration with SequeL (cf. sec. 1 ).

In these works, we develop an original (the originality fades away as years pass since this idea is exploited by other researchers.) point of view on such systems. While traditional (before say 2010) recommendation systems were seen as solving a supervised learning task, or a ranking task, we have developed the idea that recommender systems are really a problem of sequential decision making under uncertainty.

We also started a new work aiming to introduce deep learning in recommender systems. An engineer (Florian Strub) was recruited to work on this topic and presented some results at the NIPS'2015 workshop on “Machine Learning for (e-)Commerce”. Moreover we released some code to handle sparse data with the Torch7 framework and GPUs https://github.com/fstrub95/nnsparse .