Why User Control in Recommender Systems?

The theme of my RecSys 2015 paper, along with the other papers in its session, is on giving users control over their recommendation experience.

Why do we want to do this? Isn't the idea of recommender systems to figure out what the user wants and give it to them, without needing significant intervention?

There are a few reasons I think user control is an important research direction for recommender systems. First, different users have different needs, and different algorithms have different strengths. This is the idea behind McNee's human-recommender interaction framework, and the thesis and results of several of my experiments. So far, we don't have good meta-recommenders for identifying which recommender will best meet a particular user's needs, so giving them control is a way to punt on this.

First-and-a-half, if we give users control in the short term, then we can obtain more training data to develop potential meta-recommenders to provide a better user experience.

Second, transparency and control may promote trust in the recommender system. Users have a wide range of feelings and opinions about recommender systems, including in some cases fear of losing control or of being ‘bubbled’. These fears may or may not be well-grounded, but even if they aren't, the concern itself is very real. Providing users with insight into the recommendation process and the ability to manipulate it may help them feel that they can trust the system. This thesis needs to be empirically validated, but I think it's a sound working hypothesis for conducting a stream of research and doing some system design.

Third, some users really like control, and providing it can cater to such users' stated desires. Sometimes we need to give users what they need or want, but sometimes it's good to give them what they say they want.

So there's a few reasons for pursuing this line of research. I'm happy to talk more about it at RecSys — just hunt me down in one of the breaks.