Soar - Consistent Environments

Description

• Soar's learning mechanism, chunking, is a reflexive mechanism . Thus, Soar cannot help but learn new rules as it acts.
• In environments with consistent laws of behavior and relationships among facts to be learned this is an efficient means of gathering knowledge.
• However, in environments where the rules are inconsistent then this can be detrimental to Soar's performance. In these instances, Soar will generate incorrect productions which will clutter up its long-term memory and degrade performance.