Recommendations in the real world

Off-the-shelf recommenders are important building blocks for developing personalised recommendation systems, but they are not sufficient to solve personalisation problems by themselves. Such systems are designed to use either explicit or implicit data, but never both. They are unable to identify anomalous user activity, and cannot deal with new products or changing tastes.

This talk will look at the pain points of recommendation algorithms and cover ways to overcome them in practise. We will dive into the example of data drift in recommendation using data from a music streaming service. We’ll talk about why users’ tastes might change, how to identify when tastes have changed, show how to find content which will delight our users without reopening old wounds, and give some simple suggestions for how to incorporate these findings into off-the-shelf recommenders to give a more robust user experience.