ECIR 2012 took place in Barcelona at a campus of the Pompeu Fabra University which is situated not far from the city centre. Despite the size of the city, most ECIR participants stayed around the same area and wherever you went, you inadvertently bumped into a few other fellow attendees every day & night.
The conference started off with the tutorials and workshops on Sunday, though at 9:30am a lot of people still looked a bit bleary-eyed. It proved to be a popular day, a large number of people attended and the workshop organizers did their best to make it a fun and interactive event.
Monday morning, the conference itself began. Paolo Boldi gave a very interesting and enthusiastic keynote about graph centrality measures and we all learnt that Lin centrality and Paolo’s harmonic centrality are the way forward! In contrast to a few previous editions, parallel sessions had been abandoned and all sessions were held in the same room. I very much appreciated the wider variety of topics one is exposed to this way, though some sessions suffered from it (the system efficiency session being the main casualty with very few attendees scattered across the room).
The poster and demo session on Tuesday evening was a big success with lots of works, lots of good food & wine and generally a great atmosphere. Many people stayed around for a long time and the poster & demo presenters were always busy pitching their idea to someone.
One of the Wednesday sessions contained an interesting clash of industry vs. academia. An industry researcher challenged an academic presenter on his use of a publicly available data set which apparently is not representative of today’s user behavior and quite different from the data the industry researcher had collected from millions of users. The suggestion to the academic presenter to “just create your own data set” (somehow, it would have been great to be given a hint of how to do that exactly) was not well received by the audience and the obvious question from a member of the audience to the industry researcher whether their million user data set would be released was met with a round of applause. The reply was the usual one: … AOL query log… cannot release … no. This is certainly a discussion that will not go away in the near future.
Overall, ECIR 2012 was well worth a visit and the organizers did a great job of trying to accommodate everyone – a final case in point being the big screen organized at the banquet venue to keep those football fans happy that did not manage to score tickets to the Barcelona-Milan game.