WEDNESDAY, AUGUST 10, 2005
ICML and ILP 2005

This zear I had a paper accepted at the Learning Language in Logic Workshop for ICML, which was co-locacted with ILP in Bonn, Germany. I'd been to ICML in 2003 and ILP in 2004 so I knew what to expect for each conference; ICML was a hodgepodge of current Machine Learning research, with more and more of what seemed to be Kernel Methods, Bayesian Statistics, Gausians and Dirschlets, but also a fair amount on structured data learning, while ILP was definitely focused on learning in relational domains, some people working with Prolog and others just with interesting domains. I headed out on August 4th, and met up with Jesse, David Page, Soumya, Jerry and Hector, quicklz forming the Wisconsin gang. I was suprized how easily I could adapt to the local way of doing things, such as always waiting for the light to turn green before crossing the street, and also by how many people spoke english if you asked them politely. Even without english I could get around and buy food and games just fine.

It was neat to see all the ILP celeberities I had met last year and hear them talk, some of them recognized me and at least one person was using our Prolog version of the biomedical information extraction dataset, so that was very encouraging. We found a fair number of places to eat, visiting sandwich shops and pizza places, and were treated to wonderful conference dinners both with an accompanying boat ride on the Rhine; it's an amazingly scenic area of Germany that we're visiting. I found some time to visit a few game stores, finding great deals on Goldland, Dshunke, and a few of the small Adlung card games. There were a number of good papers that I'll need to read from the conference, and I was able to influence a few people to start using recall_precision curves and our dataset, so our work might get some more exposure in the future. Everyone from our group gave an effective talk, something I think has a lot to do with the "practice talk" environment we have at Wisconsin, you give your talk and we simulate conference questions, as well as give advice on each slide. I'm not sure other universities have as close of a ML department as we do. On Friday afternoon, Laura arrived at the train station, which I'm sure she'll tell you more about in her entry. We're off to Köln and Koblenz tomorrow, so we'll try to update the blog more as we can, see you back in the states in about two weeks!

Posted by Mark @ 12:15 PM CDT [Previous] [Next]