Accounting for Similarity-Based Reasoning within a Cognitive Architecture

This work explores the importance of similarity-based processes in human everyday reasoning, beyond purely rule-based processes prevalent in AI and cognitive science. A unified framework encompassing both rule based and similarity-based reasoning may provide explanations for a variety of human reasoning data. The paper implements this analysis in a cognitive architecture CLARION, which has previously succeeded in capturing a variety of human learning data in simulations. The exploration of similarity-based reasoning in this architecture leads to a more complete and more comprehensive framework of human reasoning and learning. The simulation within this architecture accurately captures human reasoning data, including numerical measures and verbal protocols. This work demonstrates the significant role played by similarity-based reasoning. Furthermore, it demonstrates how such a reasoning process falls out of the existing structure in the cognitive architecture CLARION.