Are Random Representations Accurate Approximations of Lexical Semantics?


A common assumption made by cognitive models is that lexical semantics can be approximated using randomly generated representations to stand in for word meaning. However, the use of random representations contains the hidden assumption that semantic similarity across randomly selected words is normally distributed. We evaluated this assumption by computing similarity distributions for randomly selected words from a number of well-know semantic measures and comparing them with the distributions from random representations commonly used in memory models.

Back to Table of Contents