Peopl e do not experience the world solely as an ordered \n sequence of events. The timing of our perc eptions and behaviors\n has as much of an imp act on our experiences as the nature of the\n events themselves. Yet many of the representation s currently\n used to model human behavior do not incorporate explicit models\n of the tem poral expression of these stimuli or actions. Dyna mic\n behavior is often modeled sequentially in such a way that its\n temporal resolution is reduced and potential nonstationarity is\n ignored for the sake of computational efficiency (as in Markov\n state-based models of behavio r), and/or causal mappings between\n observat ions and behavior are simplified to mitigate the\n sparseness of available datasets. Given that any artificial\n agent designed to interact with people will be dealing with\n intelligen t partners with rich mental representations of tim e,\n are we using the appropriate representat ions?\n
\nThis symposium is orie nted towards several different groups\n of re searchers, including, but not limited to: computer \n scientists who use machine learning techni ques to model human\n behavior, psychologists and neuroscientists who study social\n behav ior, and designers of robots or computational arti facts\n that interact naturally with humans i n real time. By bringing\n together members o f these communities through a shared interest\n in temporal representations, our goal is to ide ntify critical\n areas of study and promising techniques.\n
\n \n \nF or more information, see\n http://asimov.usc.edu/~mower/aaai10ss_time/\n
\n\n Papers on any aspect of modeling or studying the temporal\n aspects of human or human-machin e social interaction are\n welcome, includi ng reports on experimental results,\n descr iptions of implemented systems, and position paper s.\n Submission deadline is October 2, 2009 \n
\n