Integration of statistical methods in a knowledge based environment for assisting discrete simulation

The Knowledge Based System approach is proposed to interactively assist the simulation user in selecting appropriate statistical methods, applying them in the best way, and interpreting the results. The system aims to gather simulation statistical techniques scattered in the literature into the same knowledge base, and to complete this with empirical knowledge coming from experts. The authors present an overview of KB3S (Knowledge Based System for Simulation Statistics). This prototype results from the integration of several statistical modules in a knowledge-based environment. It is written using the expert system building tool KES.