Yesterday I attended a poster session at GAMES 2008, and today went downtown to AAAI 2008 to meet up with colleagues and attend the poster session over there.
So, about poster sessions. I am reminded that correctly treated, they can be one of the fastest ways of learning about someone's research, since they provide the opportunity to quiz the presenter about any aspect that is unclear to you. At talks, it is usual to fail to take in some crucial aspect of the mathematical model being analysed. At a poster session, the problem can be eliminated, if you don't mind interrupting the presenter frequently. At GAMES 2008 I took in a nice poster presentation by Liad Wagman (Vincent Conitzer was co-author) on "Optimal False-Name-Proof Voting Rules with Costly Voting", which happened to also be a paper at AAAI that picked up an Outstanding Paper award (I missed the talk because the room was packed out). I think the other 2 posters at GAMES that I took in in detail were "The Theory of Collusion under Financial Constraints" (Yosuke Yasuda) and "On loss aversion in a bargaining game with alternating offers" (Bram Driesen). At AAAI, where the subject matter was, shall we say, highly diversified, I found myself gravitating towards the posters on graphics and image processing; there was one on rendering scenes based on natural-language descriptions, another on scribble analysis - given a scribble over an element of a sketch, is it a filling-in or a deletion?, another on domino tiling (given a fixed collection of dominos, arrange them to form a best-possible picture of a target image. More relevant to my own research, I take in a poster by Piotr Faliszewski, about how, despite the NP-hardness of manipulating elections, there is an approximation algorithm.
At GAMES, I attended an entertaining performance by stand-up economist Yoram Bauman.