Random walks on semantic networks can resemble optimal foraging.

Abstract

When people are asked to retrieve members of a category from memory, clusters of semantically related items tend to be retrieved together. A recent article by Hills, Jones, and Todd (2012) argued that this pattern reflects a process similar to optimal strategies for foraging for food in patchy spatial environments, with an individual making a strategic decision to switch away from a cluster of related information as it becomes depleted. We demonstrate that similar behavioral phenomena also emerge from a random walk on a semantic network derived from human word-association data. Random walks provide an alternative account of how people search their memories, postulating an undirected rather than a strategic search process. We show that results resembling optimal foraging are produced by random walks when related items are close together in the semantic network. These findings are reminiscent of arguments from the debate on mental imagery, showing how different processes can produce similar results when operating on different representations.

Estimating Semantic Networks of Groups and Individuals from Fluency Data

“…Traditionally, semantic fluency data has been explained as the result of a two-stage retrieval process, where participants switch between a global cue (e.g., animals) and local cue (e.g., pets) in a spatial representation when searching for category members (Hills et al 2012). However, it has also been shown that semantic fluency data can be explained by a single-stage retrieval process on a different semantic representation: a censored random walk on a semantic network (Abbott et al 2015). Under this model, fluency data is generated by traversing edges at random in a semantic network.…”

“…Recent work in psychology has identified several potential process models (Abbott et al 2015;Hills et al 2012;Zemla and Austerweil 2017) that generate clustered fluency data. We propose a technique that capitalizes on these advances by inverting a process model used to simulate fluency data-specifically, the censored random walk model of search on a semantic network (Abbott et al 2015).…”

“…We propose a technique that capitalizes on these advances by inverting a process model used to simulate fluency data-specifically, the censored random walk model of search on a semantic network (Abbott et al 2015).…”

“…The semantic networks used by Abbott et al (2015) were pre-defined by the researchers (they showed two sensible networks both produced patchy behavior). How should a researcher choose one semantic network or another?…”

“…One process that generates clustered responses and mimics semantic fluency data is a random walk on a semantic network (Abbott et al 2015;Goñi et al 2010;Zemla and Austerweil 2017). In this article, we develop and validate a novel method for estimating semantic networks that reflects this process.…”