Exploring Integrated Co-occurrence and Semantic Mechanisms for Long Term Memory Retrieval

Semantic and co-occurrence memory associations aid the re- trieval of relevant memory elements from long term memory, but little is understood about how semantics and co-occurrence interact in this process. This paper explores the relationship between these associations via computational memory modeling in a Bayesian framework. We assessed the performance of eleven candidate mechanisms on two linguistic tasks - the Word Sense Disambiguation task and the Remote Associates Test. The most successful mechanisms use co-occurrence associations to modulate semantic associations by removing from or adding to the context or pool of candidates for retrieval, consistent with recent experimental work in memory retrieval. Although these results are a promising first step for understanding the relationship between semantic and co-occurrence associations in memory retrieval, more empirical human data is needed to validate the proposed interactions between these associations.

Recommended citation: Gebhart, L., & Li, J. (2025). " Exploring Integrated Co-occurrence and Semantic Mechanisms for Long Term Memory Retrieval." International Conference on Cognitive Modeling.
Download Paper