News

Dec 12, 2021: Text Detoxification Based on Parallel Corpora

Identification of toxicity in user texts is an active area of research. While algorithms in social networks and chats automatically identify and block comments with toxic speech, we suggest a proactive reaction to toxicity from the user. Namely, we aim at suggesting a neutral version of a user message which preserves meaningful content.

We invite you to participate in the shared task on text detoxification for the Russian language:

Dec 15, 2019: A Shared Task on Taxonomy Enrichment for the Russian Language

Taxonomies are tree structures which organize terms into a semantic hierarchy. Taxonomic relations (or hypernyms) are “is-a” relations: cat is-a animal, banana is-a fruit, Microsoft is-a company, etc. This type of relations is useful in a wide range of natural language processing tasks for performing semantic analysis. The goal of this semantic task is to extend an existing taxonomy with relations of previously unseen words.

We invite you to participate in the shared task on taxonomy enrichment for the Russian language:

Oct 30, 2018: Human-Annotated Sense-Disambiguated Word Contexts for Russian

We release a subset of the bts-rnc dataset which was annotated using crowdsourcing. The dataset is available on Zenodo via doi:10.5281/zenodo.1117228.

DOI

This dataset contains human-annotated sense identifiers for 2562 contexts of 20 words used in the RUSSE’2018 shared task on Word Sense Induction and Disambiguation for the Russian language. These sense identifiers are disambiguated as according to the sense inventory of the Large Explanatory Dictionary of Russian.

The organizers’ paper is published in the proceedings of Dialogue 2018. Please cite it as follows:

@inproceedings{Panchenko:18:dialogue,
  author    = {Panchenko, Alexander and Lopukhina, Anastasia and Ustalov, Dmitry and Lopukhin, Konstantin and Arefyev, Nikolay and Leontyev, Alexey and Loukachevitch, Natalia},
  title     = {{RUSSE'2018: A Shared Task on Word Sense Induction for the Russian Language}},
  booktitle = {Computational Linguistics and Intellectual Technologies: Papers from the Annual International Conference ``Dialogue''},
  year      = {2018},
  pages     = {547--564},
  url       = {http://www.dialog-21.ru/media/4539/panchenkoaplusetal.pdf},
  address   = {Moscow, Russia},
  publisher = {RSUH},
  issn      = {2221-7932},
  language  = {english},
}

Jun 2, 2018: RUSSE at Dialogue 2018

On May 31, 2018, the workshop on the RUSSE’2018 results was held on the Dialogue 2018 conference. The participants presented the details of their approaches and the organizers summarized the successful results of the first shared task on word sense induction and disambiguation for the Russian language.

The organizers’ paper is published in the proceedings of Dialogue 2018. Please cite it as follows:

@inproceedings{Panchenko:18:dialogue,
  author    = {Panchenko, Alexander and Lopukhina, Anastasia and Ustalov, Dmitry and Lopukhin, Konstantin and Arefyev, Nikolay and Leontyev, Alexey and Loukachevitch, Natalia},
  title     = {{RUSSE'2018: A Shared Task on Word Sense Induction for the Russian Language}},
  booktitle = {Computational Linguistics and Intellectual Technologies: Papers from the Annual International Conference ``Dialogue''},
  year      = {2018},
  pages     = {547--564},
  url       = {http://www.dialog-21.ru/media/4539/panchenkoaplusetal.pdf},
  address   = {Moscow, Russia},
  publisher = {RSUH},
  issn      = {2221-7932},
  language  = {english},
}

Mar 19, 2018: Preprint for the RUSSE'2018 WSI&D Task

We published a preprint for our upcoming Dialogue 2018 paper describing the recently conducted RUSSE shared task on word sense induction and disambiguation.

Update. The final version of the RUSSE’2018 paper is out.

Feb 15, 2018: Final Results Now Available

The final results of the RUSSE’2018 evaluation campaign are now available. They are available on CodaLab in the ‘Results’ section. Please use the following links to access the results per datasets:

We would like to thank everyone who participated and supported the shared task.

Dec 15, 2017: Test Datasets Release for WSI&D

We are glad to announce that the test datasets for the 2018 RUSSE Word Sense Induction and Disambiguation shared task have been released!

Note that we are using CodaLab for accepting the submissions. You need a single login on CodaLab to access our competitions on three datasets:

The participant’s kit has been updated as well.

Nov 10, 2017: ACL SIGSLAV Supports the Shared Task

We are excited to announce that our word sense induction and disambiguation task has been sponsored by SIGSLAV, the ACL Special Interest Group on Slavic Natural Language Processing.

Association for Computational Linguistics

Oct 25, 2017: ABBYY Supports the Shared Task

We are pleased to announce a new collaboration with the ABBYY company those experts will help in preparation of the datasets for the shared task word sense induction and disambiguation.

Oct 15, 2017: A Shared Task on Word Sense Induction and Disambiguation for the Russian Language

The second edition of RUSSE is announced. The task is dedicated to Word Sense Induction (WSI) is the process of automatic identification of the word senses.

Apr 10, 2016: Datasets on Russian Semantic Relatedness

We published several open language resources for semantic relatedness in Russian language.

Feb 1, 2015: Final Results Now Available

The final results of the RUSSE evaluation campaign are now available. Raw results are available here. We are looking forward meeting you at the special session of the Dialogue 2015 conference dedicated to the RUSSE shared task.

Jan 15, 2015: Deadline Extension

The submission deadline is prolonged until Tuesday, the 20th of January 2015 23.59 Moscow time.

Jan 5, 2015: Test Dataset is Now Available

Test dataset and submission form are available now. Please make a submission before the 15th of January 2015. You may re-submit your results. We will only consider the latest submission. We will manually check your submission and confirm its validity by email by the 15th of January 2015. If you received no confirmation from us by this date please send us copy of your submission to alexander.panchenko@uclouvain.be.

Dec 12, 2014: 52 Participants Registered

We are glad to inform you that now 52 participants from Belarus, Belgium, Czech Republic, Finland, Germany, Norway, Pakistan, Republic of Singapore, Russia, UK and US registered for the shared task. We would like to thank you for your interest to the evaluation campaign and wish you good results and inspiration during your submission.

We hope this evaluation campaign will be useful in your research and developments. Should you have any questions or suggestions regarding the competition please do not hesitate to contact us.

Dec 11, 2014: Facebook Page

We have now our Facebook group. We invite you to join it in order to follow updates regarding the competition.

Dec 7, 2014: News Page

Add this page to bookmarks and never miss any RUSSE announcements.

Dec 5, 2014: Training Dataset Update

Here you are a couple of updates.

  1. The task description page was updated as follows:
    • Duplicates were removed from the file “ae2-train.csv”.
    • Some clarifications regarding negative training samples and the association task were added.
    • Hyperlink to RuWaC was fixed.
    • Several new baseline systems were added, such as JWKTL.
  2. Submission form is now available at http://russe.nlpub.ru/submission/.

Nov 23, 2014: Training Data are Available

We are delighted to inform you that the task description and the training datasets are available now. Please do not hesitate to contact us if you have any questions related to the task. Feel free to address us in English or Russian languages.

The tentative schedule of the important dates is below:

  • 5th January 2015 — release of the test data (submissions should be done on the test data).
  • 15th January 2015 — deadline for submissions.
  • 25th January 2015 — announcement of evaluation results.
  • 15th February 2015 — deadline for Dialogue paper submission.

Let the evaluation begin!

Also, we have an RSS feed.

Share