Building knowledge-based expert systems: a tutorial
Authors: Alberto Piatti, Alessandro Antonucci and Marco Zaffalon
Abstract: Knowledge-based systems are computer programs achieving expert-level competence in solving problems for specific task areas. This chapter is a tutorial on the construction of knowledge-based systems in the theoretical framework of credal networks. Credal networks are a generalization of Bayesian networks where credal sets, i.e., closed convex sets of probability measures, are used instead of precise probabilities. This allows for a more flexible process of elicitation than in the case of Bayesian networks. In fact, credal sets allow to represent ambiguity, contrast and contradiction in a natural and realistic way. The procedure we propose is based on a sharp distinction between the domain knowledge and the process linking this knowledge to the perceived evidence, which we call the observational process. This distinction leads to a very flexible representation of both domain knowledge and knowledge about the way the information is collected, together with a procedure of aggregation of the information coming from the different sources. The overall procedure is illustrated along the chapter by a simple knowledge-based system for the prediction of the result of a football match.
Details: In Baswell, A.R. (Eds.), Advances in Mathematics Research 11, Nova Science Publishers, New York.
A version similar to the published paper can be downloaded.