Features of word similarity
This work addresses the problem of understanding word similarity for computational linguistics and psychology, but it is incremental as it builds on existing models and highlights limitations without major breakthroughs.
The study compared 28 computational models of word similarity using regression and predictive modeling on 900 rating data points, finding that word similarity ratings involve more than just semantic relatedness, with models showing limited cross-validated performance.
In this theoretical note we compare different types of computational models of word similarity and association in their ability to predict a set of about 900 rating data. Using regression and predictive modeling tools (neural net, decision tree) the performance of a total of 28 models using different combinations of both surface and semantic word features is evaluated. The results present evidence for the hypothesis that word similarity ratings are based on more than only semantic relatedness. The limited cross-validated performance of the models asks for the development of psychological process models of the word similarity rating task.