Using IBM Watson to build a podcast annotator

Some time ago now – back in October – I joined Alvaro Martinez Barrio and Mikael Huss in the Uppsala Hackup Hackathon. We formed the formidable team the “hungry aardvarks” (a name graciously provided by the organizers which was to cool to change). The goal of the hackathon was to play around with the IBM Watson APIs, and it was tons of fun. Since then I’ve been keeping busy with other things – but now I’ve finally gotten around to cleaning the code up a bit and placing it on github.

While early on we were trying to come up with some idea for something to do in bioinformatics using Watson – we realized that most of the APIs available lent themselves better to do something based on machine understanding of natural language. Mikael came up with the idea to do an automatic podcast annotator. This sounded like a really cool project and something that we could hack together in a day – so that’s what we decided to do.

You find the code here:

Our idea and execution took us in the the final – but in the end the win was snatched by the team building a speech enabled home-automation system on a raspberry pi. Super cool project – well worth the victory!

Mikael has already written about Hackup here. Go and check it out! He writes more about the our process and what the other teams at the hackathon were up to.

Finally lot’s of kudos to the organized of the Hackaton – Jason Dainter, and to IBM for build such a cool system and making it easy to use.