ABSent: Cross-Lingual Sentence Representation Mapping with Bidirectional GANs


  • Zuohui Fu Rutgers University
  • Yikun Xian Rutgers University
  • Shijie Geng Rutgers University
  • Yingqiang Ge Rutgers University
  • Yuting Wang Rutgers University
  • Xin Dong Rutgers University
  • Guang Wang Rutgers University
  • Gerard de Melo Rutgers University




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. The experiments show that our method outperforms several technically more powerful approaches, especially under challenging low-resource circumstances. The source code is available from https://github.com/zuohuif/ABSent along with relevant datasets.




How to Cite

Fu, Z., Xian, Y., Geng, S., Ge, Y., Wang, Y., Dong, X., Wang, G., & de Melo, G. (2020). ABSent: Cross-Lingual Sentence Representation Mapping with Bidirectional GANs. Proceedings of the AAAI Conference on Artificial Intelligence, 34(05), 7756-7763. https://doi.org/10.1609/aaai.v34i05.6279



AAAI Technical Track: Natural Language Processing