PITCHER

PITCHER is an Expert System I wrote using CLIPS & the Java Expert Systems Shell (http://herzberg.ca.sandia.gov). It asks the user a series of questions relating to a baseball situation, and then it provides the user with the pitch it thinks is optimal for the pitcher to throw.

The main features I added to this project were:

• robust rule system for decision-making process
• handling of non-standard input

The most difficult part was the knowledge engineering. I interviewed my high school baseball coach and asked him what he would do in many situations, and after taking notes on his decision-making process, I managed to convert his knowledge into concrete rules. Rules asserted facts with integral values, allowing for many comparisons using standard Java operators (<, >, =).

To run:

Run in a LISP environment with an inference engine (such as CLIPS):

(batch PITCHER.clp)