A critique of word similarity as a method for evaluating distributional semantic models - View it on GitHub