ACT-R (formally known as ACT*) has been developed by Anderson, John Robert and others over the past 40 years. It comprises a large long- term memory in the form of a semantic net, a small working memory of active items and a production system that operates on the memories. ACT-R includes mechanisms (i) to make commonly used instances of general rules more specific (called proceduralisation) and also to take multiple specifis rules with similar structure and create a single more abstract rule (called generalisation).
Defined on pages 536, 536, 536
Used on pages 536, 537, 549, 551
Also known as proceduralization, generalization