Skip to content


Folders and files

Last commit message
Last commit date

Latest commit



8 Commits

Repository files navigation

MultiMirror: Neural Cross-lingual Word Alignment for Multilingual Word Sense Disambiguation

This repo hosts the resources released in MultiMirror: Neural Cross-lingual Word Alignment for Multilingual Word Sense Disambiguation, a paper we presented at IJCAI 2021. MultiMirror is a cross-lingual sense projection approach for multilingual WSD based on a novel discriminative word alignment model. The sense-tagged datasets it produces lead a standard WSD classifier to achieve state-of-the-art performances on established benchmarks in French, German, Italian, Spanish and Japanese.

In data/mwsd/ you can find instructions to download the generated sense-tagged corpora in the afore-mentioned languages. Furthermore, data/alignment/ also reports how to download the manually-annotated word alignments datasets we produced for French, German, Italian and Spanish.

To train the WSD models, we suggest using wsd, the library we developed for WSD classification.

If you find our resources useful in your work, please cite us with:

  title     = {MultiMirror: Neural Cross-lingual Word Alignment for Multilingual Word Sense Disambiguation},
  author    = {Procopio, Luigi and Barba, Edoardo and Martelli, Federico and Navigli, Roberto},
  booktitle = {Proceedings of the Thirtieth International Joint Conference on
               Artificial Intelligence, {IJCAI-21}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Zhi-Hua Zhou},
  pages     = {3915--3921},
  year      = {2021},
  month     = {8},
  note      = {Main Track}
  doi       = {10.24963/ijcai.2021/539},
  url       = {},


This project is released under the CC-BY-NC-SA 4.0 license (see LICENSE).


The authors gratefully acknowledge the support of the ERC Consolidator Grant MOUSSE No. 726487 and the ELEXIS project No. 731015 under the European Union’s Horizon 2020 research and innovation programme.

This work was supported in part by the MIUR under the grant "Dipartimenti di eccellenza 2018-2022" of the Department of Computer Science of the Sapienza University of Rome.


No description, website, or topics provided.






No releases published


No packages published