Preprint 10/2022

A Novel Local Centrality Measure for Contextual Diversity in Semantic Networks

Haim Cohen, Yinon Nachshon, Paz M. Naim, Jürgen Jost, Emil Saucan, and Anat Maril

Contact the author: Please use for correspondence this email.
Submission date: 08. Mar. 2022
Pages: 53
Download full preprint: PDF (4196 kB)

This paper offers a novel local centrality-based betweenness measure designed to capture, within a semantic network, the extent to which a word is characterized by contextual diversity (CD). A CD word is one that occurs in several different and distinct contexts. After presenting the measure, we demonstrate empirically that it differs from other leading central measures, such as betweenness, degree, closeness, and the number of triangles. We then examine the relationship between the CD level of a word, as determined by the novel centrality measure, and the accessibility to knowledge stored in memory. To do so, we show that CD words are significantly more effective than non-CD words in facilitating the retrieval of subsequent words. CD words themselves, however, are not retrieved significantly faster than non-CD words. These results were obtained for a serial semantic memory task, where the word’s location constitutes a significant mediator in the relationship between the proposed measure and accessibility to knowledge stored in memory. Finally, we interpret these results as a psychological validation of our proposed measure.

18.03.2022, 12:25