Associative Exemplification Method Using Generative Artificial Intelligence to Support Associative Memory in Method of Loci

Abstract

This study proposes a method using generative AI to support associative memory formation between memory targets and location cues in the Method of Loci (MoL), which is often challenging for beginners. Based on the levels-of-processing model from cognitive psychology, our system generates three types of associative examples (physical, phonological, and semantic), allowing learners to select associations that best match their cognitive style. Experimental results with 30 university students showed that 80% of participants found it easier to create associations with AI examples, and memory retention tests after 24 hours demonstrated significantly improved recall rates. Participants particularly valued the “impressive and novel perspectives” provided by AI examples. However, when AI examples significantly differed from learners’ thought patterns, they could potentially hinder memory formation. Future work should focus on analyzing learners’ cognitive patterns beforehand and incorporating this information into AI prompts to create more personalized associative examples. This research provides guidelines for utilizing generative AI in education while suggesting improvement.