This thesis discusses ways to employ a certain mathematical characterization of similarity, kernel functions, and machine learning techniques building on it to abstract from data-oriented models of language.
A prominent ...
We propose a new method for empirically determining lists of basic concepts for the purpose of compiling extensive lexicostatistical databases. The idea is to approximate a notion of “swadeshness” formally and reproducibly ...
