How to Discover a Knowledge Representation for Causal Reasoning by Studying an Expert Physician

Abstract

The ability to identify and repr ese nt the knowl edge that a human e;<pe rt has about a particular domain is a key method in the creation of expert computer system. The first part of th is paper demonstrates a methodology for collecting and e. nalyLing observations of experts at work, in orcl 8r !o find the concep tu al framework used for the particular domain. The seco nd part develops a representation for qualitative knowledge of the structure. and behavior of a m ec hanism . The qualitative simulation, or envisionment , process is given a qualitative structural description of a m ec hanism and some initialization information, and produces a detailed description of the mechanism 's behavior. This "vertical" slice of the constr uction of a cognitive model demonstrates an effective knowledge acquisition method for the purpose of determining th e structure of the representation itself, not simrly the content of the knowledge to be encoded in that repres2ntation . Most importantly, it demonstrates th e interaction amon9 constraints derivGd from the textbook knowledge of the domain, from observations of the human expert, and from the computational requirements of successful performance.