PeopleofSuperintelligence

Researcher

Fethi Bougares

Fethi Bougares is a speech and language researcher who coauthored the 2014 RNN Encoder–Decoder paper on machine translation and now works on speech technology for Arabic dialects. They completed a doctoral thesis on automatic speech transcription systems at Le Mans in 2012 and worked as an associate professor in the LIUM laboratory. Their coauthored work includes a 2017 corpus of 17,000 annotated Tunisian Facebook comments for sentiment analysis, the findings of the 2018 shared task on multimodal machine translation, and the ON-TRAC consortium's systems for IWSLT 2020, where they led the LIUM team. In 2021 they were seconded to ELYADATA to develop speech recognition for Arabic dialects. They are head of research at ELYADATA and list an affiliation with Avignon Université.

Now at
ELYADATA and Avignon University
Badges

Timeline

  1. 2012

    Doctorate on speech transcription

    They complete a doctoral thesis at Le Mans in 2012 on automatic speech transcription systems, titled Attelage de systèmes de transcription automatique de la parole. The paper "Low latency combination of parallelized single-pass LVCSR systems", written with Mickael Rouvier, Yannick Estève and Georges Linarès, appears at Interspeech in 2012.

  2. 2014

    RNN Encoder–Decoder for translation

    In 2014, with Kyunghyun Cho, Bart van Merriënboer, Caglar Gulcehre, Dzmitry Bahdanau, Holger Schwenk and Yoshua Bengio, they coauthor "Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation" for EMNLP in Doha. The paper proposes an RNN Encoder–Decoder in which one recurrent network encodes a sequence of symbols into a fixed-length vector and another decodes it, together with a new hidden unit that has reset and update gates. Bougares and Schwenk are at Université du Maine.

  3. 2015

    Monolingual data for neural translation

    In 2015, they coauthor "On Using Monolingual Corpora in Neural Machine Translation" with Caglar Gulcehre, Orhan Firat, Kelvin Xu, Kyunghyun Cho, Loïc Barrault, Huei-Chi Lin, Holger Schwenk and Yoshua Bengio. The work leverages abundant monolingual corpora for neural machine translation and gains up to 1.96 BLEU on Turkish-English and 1.59 BLEU on Chinese-English chat messages.

  4. 2017–2018

    Tunisian dialect corpus and ASR

    In 2017, with Salima Medhaffar, Yannick Estève and Lamia Hadrich-Belguith, they present a Tunisian sentiment-analysis study at the Third Arabic Natural Language Processing Workshop in Valencia, built on a freely available corpus of 17,000 annotated Facebook comments. In 2018, with Abir Masmoudi, Mariem Ellouze, Yannick Estève and Lamia Belguith, they publish "Automatic speech recognition system for Tunisian dialect" in Language Resources and Evaluation.

  5. 2018

    Multimodal translation shared task

    In 2018, at the Third Conference on Machine Translation in Brussels, they coauthor the findings of the third shared task on multimodal machine translation, with Loïc Barrault, Lucia Specia, Chiraag Lala, Desmond Elliott and Stella Frank. The task draws 45 systems from seven teams, translating English sentences supplemented by an image into German, French or Czech. Multi30k, a dataset made with Loïc Barrault and Ozan Caglayan, extends Flickr30K with German sentences.

  6. 2020

    ON-TRAC at IWSLT 2020

    In 2020, they coauthor the ON-TRAC Consortium systems for the IWSLT 2020 offline and simultaneous speech translation tracks, a joint effort of researchers from LIA at Avignon Université, LIG at Université Grenoble Alpes and LIUM at Le Mans Université. As an associate professor, they lead the LIUM team in the consortium.

  7. January 2021–July 2021

    A secondment at ELYADATA

    From January 2021 to July 2021, they are seconded from Le Mans University to ELYADATA to develop speech recognition for Arabic dialects. The secondment is part of the ESPERANTO project, which receives funding from the European Union's Horizon 2020 programme under a Marie Skłodowska-Curie grant.

  8. As of October 2026

    Head of research at ELYADATA

    They are head of research at ELYADATA and have an affiliation with Avignon Université. In November 2025, they coauthor TEDxTN, the first publicly available Tunisian Arabic to English speech translation dataset, with 108 TEDx talks and 25 hours of speech. In September 2026, they are among the authors of the NADI 2026 overview, a shared task with five tasks on Arabic speech recognition, dialect identification, speech synthesis, translation and understanding.

Career

  1. Since 2021ELYADATAHead of Research
  2. NowAvignon University
  3. From 2014University of Maine
  4. Le Mans UniversityAssociate Professor

Education

  1. 2012Le Mans UniversityDoctorate

From the biography.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  2. The United States employs an estimated 48% of people shaping AIWe estimate where the people shaping AI work today, how universities, companies and research institutes split across countries, and how country shares have changed since the 1960s.
  3. 43% of living people shaping AI work outside their country of birthWe estimate that 43% of living people shaping AI work outside their country of birth (520 of 1,198 people with known birth and employment countries).
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionWe estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Fethi Bougares yet. Send us a free-licensed photo.

  1. Fethi Bougares – Laboratoire d'Informatique de l'Université du Mans lium.univ-lemans.fr. Checked 21 October 2026.
  2. Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation - ACL Anthology aclanthology.org. Checked 21 October 2026.
  3. Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation (ar5iv) ar5iv.labs.arxiv.org. Checked 21 October 2026.
  4. On Using Monolingual Corpora in Neural Machine Translation arxiv.org. Checked 21 October 2026.
  5. Sentiment Analysis of Tunisian Dialects: Linguistic Ressources and Experiments - ACL Anthology aclanthology.org. Checked 21 October 2026.
  6. Findings of the Third Shared Task on Multimodal Machine Translation - ACL Anthology aclanthology.org. Checked 21 October 2026.
  7. Multi30k – Laboratoire d'Informatique de l'Université du Mans lium.univ-lemans.fr. Checked 21 October 2026.
  8. ON-TRAC Consortium for End-to-End and Simultaneous Speech Translation Challenge Tasks at IWSLT 2020 - ACL Anthology aclanthology.org. Checked 21 October 2026.
  9. ON-TRAC Consortium – On-TRAC on-trac.univ-avignon.fr. Checked 21 October 2026.
  10. Secondment of Fethi Bougares - Projet ESPERANTO esperanto.univ-lemans.fr. Checked 21 October 2026.
  11. fbougares (fethi Bougares) · GitHub github.com. Checked 21 October 2026.
  12. TEDxTN: A Three-way Speech Translation Corpus for Code-Switched Tunisian Arabic - English arxiv.org. Checked 21 October 2026.
  13. NADI 2026: The Second Multidialectal Arabic Speech Processing Shared Task arxiv.org. Checked 21 October 2026.