Translation Acquisition Using Synonym Sets

Daniel Andrade, Masaki Tsuchida, Takashi Onishi and Kai Ishikawa

We propose a new method for translation acquisition which uses a set of synonyms to acquire translations from comparable corpora. The motivation is that, given a certain query term, it is often possible for a user to specify one or more synonyms. Using the resulting set of query terms has the advantage that we can overcome the problem that a single query term's context vector does not always reliably represent a terms meaning due to the context vector's sparsity. Our proposed method uses a weighted average of the synonyms' context vectors, that is derived by inferring the mean vector of the von Mises-Fisher distribution. We evaluate our method, using the synsets from the cross-lingually aligned Japanese and English WordNet. The experiments show that our proposed method significantly improves translation accuracy when compared to a previous method for smoothing context vectors.