Semi-Supervised Bilingual Lexicon Induction with Two-way Interaction

TL;DR

This paper introduces a semi-supervised framework for bilingual lexicon induction that enhances performance by facilitating interaction between supervised and unsupervised data through message-passing mechanisms, demonstrating significant improvements on multiple datasets.

Contribution

It proposes a novel semi-supervised BLI framework with two message-passing mechanisms for better knowledge transfer, applicable to various optimal transport-based methods.

Findings

Significant performance improvements on MUSE and VecMap datasets.

The two-way interaction mechanism is key to performance gains.

Robustness demonstrated on distant language pairs.

Abstract

Semi-supervision is a promising paradigm for Bilingual Lexicon Induction (BLI) with limited annotations. However, previous semisupervised methods do not fully utilize the knowledge hidden in annotated and nonannotated data, which hinders further improvement of their performance. In this paper, we propose a new semi-supervised BLI framework to encourage the interaction between the supervised signal and unsupervised alignment. We design two message-passing mechanisms to transfer knowledge between annotated and non-annotated data, named prior optimal transport and bi-directional lexicon update respectively. Then, we perform semi-supervised learning based on a cyclic or a parallel parameter feeding routine to update our models. Our framework is a general framework that can incorporate any supervised and unsupervised BLI methods based on optimal transport. Experimental results on MUSE and VecMap datasets show significant improvement of our models. Ablation study also proves that the two-way interaction between the supervised signal and unsupervised alignment accounts for the gain of the overall performance. Results on distant language pairs further illustrate the advantage and robustness of our proposed method.