Thursday, April 05, 2007
ICWSM In Full
(International Conference on Weblogs and Social Media)
Following my return from my US trip, I felt it really only fair to give a more in depth account of my trip. I shall start here by giving more details on the conference. I may name names and I may say some negative things, though not necessarily relating the two. You 'll just have to wait and see.
Firstly, the positive. It was good to catch up with people who I met at last years AAAI Symposium. Obviously the organisers, Nicolas Nicolov, Natalie Glance and that Matt Hurst fellow. There were also a few attendees who I got to see again, including Richard Tong and Diego Reforgiato. There were also a number of other interesting people there, particularly the representatives of OpinMind and TEXTMAP. There were a broad range of talks, and helpfully they were grouped into sessions this year. Of most interest to me was the plethora of work on sentiment analysis. People are still taking very general approaches to this, and despite the impressiveness, there is considerable room for refinement and improvement. It was handy having the venue in the hotel, but jet lag prevented me from seeing much of Boulder.
And now, the negatives, many of which were totally out of the control of the organisers, so I hope they don't take too much of this to heart. It didn't feel like there was as much work of a sociological bent this time around, despite there being a loose track in which Sam Gosling and I both fell. Following this, it was a shame not to see people like Shlomo Argamon there again. It is also true to say that there were some pretty poor presentations. Apart from an interesting study confirming a belief I have always held that you should never just read your slides, many were just painful to listen to. I understand that a lot of people don't have the confidence for public speaking, but I wish there was some way I could help. Yes, talking to a room full of people is hard, but what each speaker needs to bear in mind is that for that moment, they have a room full of people dedicated to listening to them, the expert in the field. Every speaker is an expert in what they are about to say. I believe if more people realised that, there would be more life in their oration and the overall standards of talks would be considerably higher. But that's just me, I evaluate style and delivery as much, if not more, than content.
Despite the apparent lacking in attendees and confidence, one thing that was missing from the conference itself was room for discussion. AAAI had some good panel sessions and slots just for talking as a whole group. These parts were very interesting - they are an opportunity to go beyond any specific piece of research. Probably due to the volume of interest in this conference, there was no time for anything but talks. I think next time at least one good long panel session, and more time for general discussion would be really good. From conversations with the organisers, however, I think they are aware of this.
Ultimately, for me personally, the conference was a successful. Despite the presentation, it was interesting to see what the world's blogademics are up to. The best part was that it was fantastic opportunity to meet some bright people with great ideas, and I took it to the max. I bet you can't wait to read about the second leg of my trip, heading south to UT Austin.