The utility of expert knowledge

Jonathan Schacffer, T.A. Marsland · International Joint Conference on Artificial Intelligence · 1985

How useful is the knowledge we add to an expert system? What is important knowledge? Can too much knowledge be bad? These questions are examined by presenting the preliminary results of experiments that paired programs with varying amounts of chess knowledge against each other. The experiments illustrate problems of interacting knowledge and give some insight into methodologies for teaching expert systems.

Read the paper · More papers on PaperTik