Surafel Melaku Lakew

Fondazione Brunno Kessler
  • Multilingual NMT
  • Low-Resource Languages NMT
  • Multimodal NMT

I am a Ph.D. candidate at FBK-Fondazione Bruno Kessler (HLT-MT research group) and at the University of Trento (ICT-Doctoral School), Trento, Italy, working on the topic “Deep Learning for human-in-the-loop advanced Machine Translation”.

After obtaining my bachelor’s degree I have worked as a graduate assistant, at Bahir Dar University, Ethiopia. Following, I have received my M.Sc. in computer science from the Ca’Foscari University of Venice under the supervision of Prof. Marcello Pelillo and Prof. Andrea Prati (please check out my thesis work here) and an MBA from Bologna Business School in Business Management/Green Management and Sustainable Businesses.

My current research focuses on Multilingual Neural Machine Translation, particularly for low-resource languages. I am advised by Prof. Marcello Federico. Please check out some of the work from our lab below!

Curriculum Vitae
Contacts
  • Surafel M. Lakew
Links

Publications

CONFERENCE PAPERS

S. M. Lakew, A. Aerofeeva, M. Federico, “Neural Machine Translation into Language Varieties”, Proceedings of the Third Conference on Machine Translation, Brussels, Belgium, 2018.

S. M. Lakew, A. Erofeeva, M. Negri, M. Federico, M. Turchi, “Transfer Learning in Multilingual Neural Machine Translation with Dynamic Vocabulary”. Proceedings of the 15th International  Workshop on Spoken Language Translation, Bruges, Belgium, 2018.

S. M. Lakew, M. Federico,  “Adapting Multilingual NMT to Extremely Low Resource Languages, FBK’s Participation in the Basque-English Low Resource MT Task, IWSLT 2018”, Proceedings of the 15th International Workshop on Spoken Language Translation, Bruges, Belgium, 2018.

S. M. Lakew, M. Cettolo,  M. Federico, “A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translation“. 5th Italian Conference on Computational Linguistics (CLiC-it), Torino, Italy, 2018. (– to appear, in the research communications track)

S. M. Lakew, M. Cettolo, M. Federico, “A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translation“. Proceedings of the 27th International Conference on Computational Linguistics. Santa Fe, New Mexico, USA, 2018.

S. M. Lakew, M. Federico, M. Negri, M. Turchi, “Multilingual Neural Machine Translation for Low Resource Languages”. The 1st Italian Journal of Computational Linguistics (IJCoL), Rome, Italy, 2018.

S. M. Lakew, Q. F. Lotito, M. Turchi, M. Negri,  M. Federico, “Improving Zero-Shot Translation of Low-Resource Languages“. Proceedings of the 14th International  Workshop on Spoken Language Translation, pp. 35-41, Tokyo, Japan, 2017. (– Best paper award)

S. M. Lakew, Q. F. Lotito, M. Turchi, M. Negri,  M. Federico,  “FBK’s Multilingual Neural Machine Translation System for IWSLT 2017“.; Proceedings of the 14th International Workshop on Spoken Language Translation, pp. 113-119, Tokyo, Japan, 2017.

S. M. Lakew, M. A. Di Gangi, M. Federico, “Multilingual neural machine translation for low resource languages”. Proceedings of the 4th Italian Conference on Computational Linguistics (CLIC-IT), Rome, Dec 2017.

K. H. Ahmed, S. M. Lakew, M. Pelillo, A. Prati, “Using Dominant Set for Data Association in Multi-Camera Tracking“. ICDSC: Proceedings of the 2015 international conference on distributed smart cameras. 2015. pp. 38-43.

 

CONFERENCE PRESENTATIONS

Improving Zero-Shot Translation of Low-Resource Languages: 4th International Workshop on Spoken Language Translation (IWSLT 2017), Tokyo, Japan,  Dec 2017.

Multilingual Neural Machine Translation for Low Resource Languages: 4th Italian Conference on Computational Linguistics (Clic-it 2017), Rome, Italy, Dec 2017.

A Comparison of Transformer and Recurrent Neural Networks on Multilingual Neural Machine Translation: 27th International Conference on Computational Linguistics. Santa Fe, New Mexico, USA, 2018.

 

POSTER PRESENTATIONS

  • FBK’s Multilingual Neural Machine Translation System for IWSLT 2017: 14th International Workshop on Spoken Language Translation (IWSLT 2017), Tokyo, Japan,  Dec 2017.
  • Neural Machine Translation into Language Varieties: Third Conference on Machine Translation, Brussels, Belgium, 2018.
  • Transfer Learning in Multilingual Neural Machine Translation with Dynamic Vocabulary: 15th International  Workshop on Spoken Language Translation, Bruges, Belgium, 2018.
  • Adapting Multilingual NMT to Extremely Low Resource Languages, FBK’s Participation in the Basque-English Low Resource MT Task, IWSLT 2018: 15th International Workshop on Spoken Language Translation, Bruges, Belgium, 2018.

 

COURSE WORKS

These are some of the courses/training/summer schools I attended in the Ph.D. program.

  • Machine Translation | Research Methodology | Business Opportunity Recognition | Statistical Relational AI
  • 7th Lisbon Machine Learning School (LxMLS 2017)

 

OPEN SOURCE PROJECTS

  • Coming soon…