Poster C74, Sunday, March 25, 1:00-3:00 pm, Exhibit Hall C
Searching for semantic knowledge: A vector space semantic analysis of the feature generation task
Rebecca Cutler1, Nate Klooster2, Melissa Duff1, Sean Polyn1; 1Vanderbilt University, 2University of Pennsylvania
A recent neuropsychological study found that amnesic patients with hippocampal damage (HD group; n=5) reliably produce fewer responses than healthy demographically matched comparison participants (NC group; n=15) in a semantic feature generation task (Klooster & Duff, 2015), consistent with the idea that semantic cognitive deficits are associated with hippocampal damage. Participants were presented with a target word and asked to produce as many features of that word as possible. We examined these data using a semantic vector space model to characterize semantic similarity between the target word and the response words comprising the features. Overall, HD patients generated features that were closer in semantic space to the target word, as compared to the wider range of features generated by the NC group. Both groups tended to initiate the search process with features close in semantic space to the target word, with a gradual decline in similarity to the target word over the first several responses. Adjacent features in the response sequence showed stronger similarity to each other than to non-adjacent features, suggesting that the search process follows a local trajectory in semantic space. Termination of feature generation in HD patients, but not the NC group, was preceded by a steady decline in similarity to the target word, suggesting that a patient’s search process is disrupted when the search ranges too far in semantic space from the target word. We consider these results in terms of a model in which hippocampus supports the probing of semantic memory.
Topic Area: LONG-TERM MEMORY: Semantic