sanjay@ux.acs.umn.edu (Sanjay Chatterjee) (03/18/91)
A past attempt of mine at getting references from the readers of this group was very successful. I hope this new inquiry gains some responses, too. I'd like to understand the perspectives and methodologies associated with studying how agents solve problems (specifically associated with diagnostic tasks) in environments where potentially useful knowledge for performing such tasks is distributed and not 'processed' to be available and/or utilizable by these agents. I'm really at ground level, so some of the foundational literature that might address such issues will be clearly useful. At the same time, I would be very keen to learn of - say - a review article that might summarize the current status of some of the work associated with my query. In this regard, I am keen on finding work where real-world knowledge was utilized (by way of nature of information/environments/agent behavior studied) in constructing models, or even where studies were restricted to being observational / speculative in nature, without any experimental work done. I might mention that my exposure to the literature on this topic has so far been limited; I've seen some work by Durfee/Lesser/Corkill, and a couple of older papers by Chandrashekharan, and Randall Davis. I've also seen some work from an OR/management perspective by Thomas Malone. I'm sure there's much that needs to be added to this list. So please help me out! Thanks, Sanjay sanjay@ux.acs.umn.edu