A Knowledge Scout for Discovering Medical Patterns: Methodology and System SCAMP

Knowledge scouts are software agents that autonomously synthesize knowledge of interest to a given user (target knowledge) by applying inductive database operators to a local or distributed dataset. This paper describes briefly a method and a scripting language for developing knowledge scouts, and then reports on experiments with a knowledge scout, SCAMP, for discovering patterns characterizing relationships among lifestyles, symptoms and diseases in a large medical database. Discovered patterns are presented in two forms: (1) attributional rules, which are expressions in attributional calculus, and (2) association graphs, aphically and abstractly represent relations expressed by the rules. Preliminary results indicate a high potential utility of the presented methodology for deriving useful and understandable knowledge.

• Type: Conference
• Year: 2000
• Where: FQAS
• Authors: Kenneth A. Kaufman, Ryszard S. Michalski
• Keywords: Database, FQAS 2000, Inductive Database Operators, Knowledge Scout, Target Knowledge