Auto-generating Bilingual Dictionaries
Noam Ordan, Jordi Gracia Gracia, Ilan Kernerman · Dialnet (Universidad de la Rioja) · 2017
Inferring a bilingual dictionary from two or more existing bilingual dictionaries is a non-trivial task, as seen in reports of large-scale, computationally-heavy experiments published in recent years. Early works on this have already noted that the main obstacle in such inferences stems from the fact that polysemy is not isomorphic across languages, and often a monosemous lexical item in one language can be polysemous in its corresponding transltion into another language. In this paper, we propose an experiment on tr anslation inference across dictionaries, based on a graph-based view of a collection of bilingual dictionaries. The idea is to explore the results and analyze them from a lexicographic point of view to reflect the implications that the issue of anisomorphy introduces in the task, and to illustrate its hurdles and potential benefits.