ABSent: Cross-Lingual Sentence Representation Mapping with Bidirectional GANs
This addresses a domain-specific challenge in cross-lingual NLP for scenarios with restricted parallel data, representing an incremental advance over prior word-level methods.
The paper tackled the problem of obtaining cross-lingually aligned sentence representations with limited parallel data, proposing the ABSent framework that learns mappings using bidirectional GANs.
A number of cross-lingual transfer learning approaches based on neural networks have been proposed for the case when large amounts of parallel text are at our disposal. However, in many real-world settings, the size of parallel annotated training data is restricted. Additionally, prior cross-lingual mapping research has mainly focused on the word level. This raises the question of whether such techniques can also be applied to effortlessly obtain cross-lingually aligned sentence representations. To this end, we propose an Adversarial Bi-directional Sentence Embedding Mapping (ABSent) framework, which learns mappings of cross-lingual sentence representations from limited quantities of parallel data.