Year: 2015
Type: conference-paper
Abstract: Obtaining norm scores for subjective properties of words can\nbe quite cumbersome as it requires a considerable investment\nproportional to the size of the word set. We present a method\nto predict norm scores for large word sets from a word\nassociation corpus. We use similarities between word pairs,\nderived from this corpus, to construct a semantic space.\nStarting from norm scores for a subset of the words, we\nretrieve the direction in the space that optimally reflects the\nnorm data associated with the words. This direction is used to\northogonally project all the other words in the semantic space\non, providing predictions of the words on the variable of\ninterest. In this study, we predict valence, arousal, dominance,\nage of acquisition, and concreteness and show that the\npredictions correlate strongly with the judgments of human\nraters. Furthermore, we show that our predictions are superior\nto those derived using other methods
Source: Lirias
Institutions: KU Leuven, Catholic University of America
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2403710930