Figure 1 - uploaded by Andrea Bruera
Content may be subject to copyright.
Distribution of concreteness and sensory strength ratings for the 100 verb-noun polysemic phrases. Ratings (y axis) are normalized in the range 0-1. As shown by the averages (horizontal coloured lines), concrete phrases show higher concreteness and stronger involvement of all types of sensory information.
Source publication
Polysemes are words that can have different senses depending on the context of utterance: for instance, 'newspaper' can refer to an organization (as in 'manage the newspaper') or to an object (as in 'open the newspaper'). Contrary to a large body of evidence coming from psy-cholinguistics, polysemy has been traditionally modelled in NLP by assuming...
Contexts in source publication
Context 1
... the following we will describe the stimuli selection procedure in detail. A visualization of the distributions of the ratings, directly comparing abstract and concrete senses, is displayed in Figure 1. ...
Context 2
... distributions of the resulting ratings are reported in Figure 1. As it can be seen, the largest difference between distributions for concrete/abstract senses is found for concreteness, sight and touch (in all cases p < 0.0001), followed by hearing (p = 0.0163). ...
Context 3
... Section 5.3) provides the best performance in all variables except taste (XGLM sight = 0.839, XGLM touch = 0.774, XGLM hearing = 0.837, XGLM smell = 0.672; best performance in taste by Conceptnet Numberbatch numberbatch taste = 0.725). Overall low performance in taste and smell can be explained by the fact that, as shown in Figure 1, these two sensory variables had the smallest variance overall, and tended to cluster around low valuesthus making it difficult to differentiate among values for different phrases. ...
