Zero-Shot Cross-Lingual Machine Reading Comprehension via Inter-sentence Dependency Graph


  • Liyan Xu Emory University
  • Xuchao Zhang NEC Labs America
  • Bo Zong NEC Labs America
  • Yanchi Liu NEC Labs America
  • Wei Cheng NEC Labs America
  • Jingchao Ni NEC Labs America
  • Haifeng Chen NEC Labs America
  • Liang Zhao Emory University
  • Jinho D. Choi Emory University



Speech & Natural Language Processing (SNLP), Machine Learning (ML)


We target the task of cross-lingual Machine Reading Comprehension (MRC) in the direct zero-shot setting, by incorporating syntactic features from Universal Dependencies (UD), and the key features we use are the syntactic relations within each sentence. While previous work has demonstrated effective syntax-guided MRC models, we propose to adopt the inter-sentence syntactic relations, in addition to the rudimentary intra-sentence relations, to further utilize the syntactic dependencies in the multi-sentence input of the MRC task. In our approach, we build the Inter-Sentence Dependency Graph (ISDG) connecting dependency trees to form global syntactic relations across sentences. We then propose the ISDG encoder that encodes the global dependency graph, addressing the inter-sentence relations via both one-hop and multi-hop dependency paths explicitly. Experiments on three multilingual MRC datasets (XQuAD, MLQA, TyDiQA-GoldP) show that our encoder that is only trained on English is able to improve the zero-shot performance on all 14 test sets covering 8 languages, with up to 3.8 F1 / 5.2 EM improvement on-average, and 5.2 F1 / 11.2 EM on certain languages. Further analysis shows the improvement can be attributed to the attention on the cross-linguistically consistent syntactic path. Our code is available at




How to Cite

Xu, L., Zhang, X., Zong, B., Liu, Y., Cheng, W., Ni, J., Chen, H., Zhao, L., & Choi, J. D. (2022). Zero-Shot Cross-Lingual Machine Reading Comprehension via Inter-sentence Dependency Graph. Proceedings of the AAAI Conference on Artificial Intelligence, 36(10), 11538-11546.



AAAI Technical Track on Speech and Natural Language Processing