An automatic rule induction technique for learning expert systems

Abstract

An automatic rule induction technique for Learning Expert Systems (LES) has been designed and developed. Here we present a new technique for rule induction which induces several types of rules, namely Concept Rule (CR), Modified Concept Rule (MCR), and Generalized Modified Concept Rule (GMCR) from input training instances during Knowledge Acquisition (KA). These rules draws inferences during end user interaction. This new technique finds application in establishing various laws of Artificial Intelligence like inheritance and transitivity etc.

Keywords:

• Learning expert systems
• knowledge acquisition
• conceptual clustering
• analogy
• machine learning

C.R. Categories:

• I2.3
• I2.6