Greetings all, I went to my 8th CICLING conference earlier this month, and once again found it to be a very nice event. It's quite unique among the conferences that I attend in that it's relatively small, and yet there are always very good invited speakers. This year the invited speakers were Jill Burstein, Ken Church, Dekang Lin, and Bernardo Magnini. That's a fairly impressive group of speakers, and in fact I think all of them have been invited speakers at much larger events as well.
There have now been a total of 10 CICLING conferences (each year since 2000) and I have attended eight of those. The only ones I've missed have been the ones that took place outside of Mexico City (in Korea and Israel) and I guess there is no accident in that fact - Mexico City remains an endlessly interesting place to me, and I always enjoy the ambiance there. My stated purpose in attending was to present some recent work on name discrimination (using SenseClusters of course) and I think it was well received. One of the things I was reporting on was the fact that our results are actually better now when using automatic cluster stopping (that is predicting how many identities exist in the data) than back when we would manually give the clustering algorithm the correct answer, so that seemed like fairly significant progress. My next conference destination will be NAACL in Boulder, Colorado. I'll be there for the full week since I must attend some NAACL board meetings early in the week, and then give an invited talk in a workshop on the last day (on software engineering, NLP, and science...expanding a bit on the ideas in the Empiricism is not a Matter of Faith article from last year...) I'll look forward to seeing some of you there, I hope! Cordially, Ted -- Ted Pedersen http://www.d.umn.edu/~tpederse
