Multilinguality and Language Technology

MT Team

We are a group at the MLT-Lab at DFKI focusing on neural machine translation and multilingual applications.

Machine translation is the automated process of converting input in one language into another one. We work in both text and sign language translation using deep learning techniques. Besides developing generic MT systems, we specialize machine translation to settings such as low-resourced languages, highly multilingual environments or document and dialog translation.

Demo: Presidency Translator with DFKI MT engines


Some current projects:

AVASAG

Avatar-based language assistant for automated sign translation

Period: 01.05.2020–30.04.2023, funded by BMBF.

Project Page

CORA4NLP

Co(n)textual Reasoning and Adaptation for Natural Language Processing

Period: 01.10.2020-30.09.2023, funded by BMBF.

Project website

EUC-PT

EU Council Presidency Translator
Period: 01.02.2020–31.12.2020, funded by Auswärtiges Amt.

Link

Fair Forward

Consulting services to Gesellschaft für Internationale Zusammenarbeit (GIZ) on technical aspects of AI in international cooperation including natural language processing (NLP), training data and data access for FAIR Forward – Artificial Intelligence for All.

Period: 01.05.2020–30.06.2021, GIZ Project No. 19.2010.7-003.00

Project Page

SocialWear

Socially Interactive Smart Fashion

Period: 01.05.2020–30.04.2024, funded by BMBF.

Project Page


Projects participations D&R / TR

We participate in projects from the D&R and TR groups where we contribute with machine translation and multilingual technologies:

ELRC

European Language Resource Coordination

Upcoming shared task on multilingual cultural heritage translation at WMT2021.


Projects in collaboration with University of Saarland

IDEAL

Information Density and Linguistic Encoding

Period: 01.02.2018–31.01.2022, funded by DFG.

Project Page

MMT

Multimodal Machine Translation – Convergence of multiple modes of input

Period: 01.03.2019–28.02.2021, funded by SPARC.
 

Projektseite


Selected publications

  • Dana Ruiter, Cristina España-Bonet, Josef van Genabith. Self-Induced Curriculum Learning in Self-Supervised Neural Machine Translation. In: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP-2020), November 2020.
  • Jingyi Zhang, Josef van Genabith. Translation Quality Estimation by Jointly Learning to Score and Rank. In: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP-2020), November 2020.
  • Hongfei Xu, Josef van Genabith, Deyi Xiong, Qiuhui Liu. Dynamically Adjusting Transformer Batch Size by Monitoring Gradient Direction Change.  In: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics.Annual Meeting of the Association for Computational Linguistics (ACL-2020) ACL 2020 July 5-10 virtual, Pages 3519-3524, 7/2020.  
  • Hongfei Xu, Josef van Genabith, Deyi Xiong, Quihui Liu, Jingyi Zhang. Learning Source Phrase Representations for Neural Machine Translation. In: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Annual Meeting of the Association for Computational Linguistics (ACL-2020) ACL 2020 July 5-10 virtual, Pages 386-396, 7/2020.
  • Hongfei Xu, Qiuhui Liu,Josef van Genabith, Deyi Xiong, Jingyi Zhang. Lipschitz Constrained Parameter Initialization for Deep Transformers. In: Proceedings of the 58th AnnualMeeting of the Association for Computational Linguistics. Annual Meeting of the Association for Computational Linguistics (ACL-2020) ACL 2020 July 5-10 virtual, Pages 397-402, 7/2020.
  • Jesujoba O. Alabi, Kwabena Amponsah-Kaakyire, David I. Adelani, Cristina España-Bonet. Massive vs. Curated Embeddings for Low-Resourced Languages: the Case of Yorùbá and Twi. In: 12th Language Resources and Evaluation Conference (LREC). International Conference on Language Resources and Evaluation (LREC-2020) May 12-17 Marseille France European Language Resources Association (ELRA) París 2020.
  • Marta R. Costa-jussà, Cristina España-Bonet, Pascale Fung, Noah A. Smith. Multilingual and Interlingual Semantic Representations for Natural Language Processing: A Brief Introduction. In: Computational Linguistics (CL) Special Issue: Multilingual and Interlingual Semantic Representations for Natural Language Processing Pages 1-8 MIT Press Cambridge, MA, USA 2020.
  • Hongfei Xu, Deyi Xiong, Josef van Genabith, Qiuhui Liu. Efficient Context-Aware Neural Machine Translation with Layer-Wise Weighting and Input-Aware Gating. In: Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence. International Joint Conference on Artificial Intelligence (IJCAI-2020) 29th International Joint Conference on Artificial Intelligence and 17th Pacific Rim International Conference on Artificial Intelligence IJCAI-PRICAI 2020 January 5-10 Virtual, Pages 3933-3940, 7/2020.
  • Dana Ruiter, Cristina España-Bonet, Josef van Genabith. Self-Supervised Neural Machine Translation. In: 57th Annual Meeting of the Association for Computational Linguistics. Annual Meeting of the Association for Computational Linguistics (ACL-2019) 57th July 28-August 2 Florence Italy ACL 2019.
  • Cristina España-Bonet, Ádám Csaba Varga, Alberto Barrón-Cedeño, Josef van Genabith. An Empirical Analysis of NMT-Derived Interlingual Embeddings and their Use in Parallel Sentence Identification. In: IEEE Journal of Selected Topics in Signal Processing (JSTSP) 11 8 Pages 1340-1350 IEEE 12/2017

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz