Inverted Bilingual Topic Models for Lexicon Extraction from Non-parallel Data

TL;DR

This paper introduces inverted bilingual topic models that improve lexicon extraction from non-parallel data by modeling word relationships more effectively and handling noisy seed dictionaries.

Contribution

It proposes two novel bilingual topic models that invert word and document roles and incorporate translation probabilities to better extract lexicons from non-parallel corpora.

Findings

Models outperform previous methods on real-world data

Effective in handling noisy seed dictionaries

Improves semantic capture of cross-lingual words

Abstract

Topic models have been successfully applied in lexicon extraction. However, most previous methods are limited to document-aligned data. In this paper, we try to address two challenges of applying topic models to lexicon extraction in non-parallel data: 1) hard to model the word relationship and 2) noisy seed dictionary. To solve these two challenges, we propose two new bilingual topic models to better capture the semantic information of each word while discriminating the multiple translations in a noisy seed dictionary. We extend the scope of topic models by inverting the roles of "word" and "document". In addition, to solve the problem of noise in seed dictionary, we incorporate the probability of translation selection in our models. Moreover, we also propose an effective measure to evaluate the similarity of words in different languages and select the optimal translation pairs. Experimental results using real world data demonstrate the utility and efficacy of the proposed models.