Yle data science - Areena recommendation and mixed analytics

This is a technical meeting for everyone interested in data science.

The meeting will start with a visit to studios for those who are early enough. There is a limit of 40 people for the visit. Snacks offered.

Schedule:

17:15 Visiting studios (for those who come early)

18:00 Introduction (Timo Aho)

18:10 Yle Areena recommendation (Aleksi Rossi)
How does the recommendation algorithm combine both collaborative and contextual filtering and deals with cold start problem? How to analyze A/B test results and measure "good recommendations" in Yle?

18:40 Break

19:00 Analyzing customers (Timo Aho and Toni Mikkola)
Visualizing and analyzing Yle web/mobile media usage patterns, e.g., with t-SNE content clusters. Also how to predict users' demographics based on their use history?

Welcome!

Image source: Yle