{"department":[{"_id":"820"}],"author":[{"last_name":"Igamberdiev","full_name":"Igamberdiev, Timour","first_name":"Timour"},{"last_name":"Vu","full_name":"Vu, Doan Nam Long","first_name":"Doan Nam Long"},{"last_name":"Kuennecke","full_name":"Kuennecke, Felix","first_name":"Felix"},{"first_name":"Zhuo","last_name":"Yu","full_name":"Yu, Zhuo"},{"last_name":"Holmer","full_name":"Holmer, Jannik","first_name":"Jannik"},{"last_name":"Habernal","full_name":"Habernal, Ivan","first_name":"Ivan","id":"101881"}],"year":"2024","user_id":"15504","_id":"52842","title":"DP-NMT: Scalable Differentially Private Machine Translation","language":[{"iso":"eng"}],"publisher":"Association for Computational Linguistics","editor":[{"first_name":"Nikolaos","full_name":"Aletras, Nikolaos","last_name":"Aletras"},{"full_name":"De Clercq, Orphee","last_name":"De Clercq","first_name":"Orphee"}],"citation":{"ama":"Igamberdiev T, Vu DNL, Kuennecke F, Yu Z, Holmer J, Habernal I. DP-NMT: Scalable Differentially Private Machine Translation. In: Aletras N, De Clercq O, eds. <i>Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations</i>. Association for Computational Linguistics; 2024:94–105.","ieee":"T. Igamberdiev, D. N. L. Vu, F. Kuennecke, Z. Yu, J. Holmer, and I. Habernal, “DP-NMT: Scalable Differentially Private Machine Translation,” in <i>Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations</i>, 2024, pp. 94–105.","chicago":"Igamberdiev, Timour, Doan Nam Long Vu, Felix Kuennecke, Zhuo Yu, Jannik Holmer, and Ivan Habernal. “DP-NMT: Scalable Differentially Private Machine Translation.” In <i>Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations</i>, edited by Nikolaos Aletras and Orphee De Clercq, 94–105. St. Julians, Malta: Association for Computational Linguistics, 2024.","apa":"Igamberdiev, T., Vu, D. N. L., Kuennecke, F., Yu, Z., Holmer, J., &#38; Habernal, I. (2024). DP-NMT: Scalable Differentially Private Machine Translation. In N. Aletras &#38; O. De Clercq (Eds.), <i>Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations</i> (pp. 94–105). Association for Computational Linguistics.","short":"T. Igamberdiev, D.N.L. Vu, F. Kuennecke, Z. Yu, J. Holmer, I. Habernal, in: N. Aletras, O. De Clercq (Eds.), Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, Association for Computational Linguistics, St. Julians, Malta, 2024, pp. 94–105.","mla":"Igamberdiev, Timour, et al. “DP-NMT: Scalable Differentially Private Machine Translation.” <i>Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations</i>, edited by Nikolaos Aletras and Orphee De Clercq, Association for Computational Linguistics, 2024, pp. 94–105.","bibtex":"@inproceedings{Igamberdiev_Vu_Kuennecke_Yu_Holmer_Habernal_2024, place={St. Julians, Malta}, title={DP-NMT: Scalable Differentially Private Machine Translation}, booktitle={Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations}, publisher={Association for Computational Linguistics}, author={Igamberdiev, Timour and Vu, Doan Nam Long and Kuennecke, Felix and Yu, Zhuo and Holmer, Jannik and Habernal, Ivan}, editor={Aletras, Nikolaos and De Clercq, Orphee}, year={2024}, pages={94–105} }"},"type":"conference","date_updated":"2024-03-25T11:31:12Z","page":"94–105","publication":"Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations","place":"St. Julians, Malta","status":"public","date_created":"2024-03-25T11:30:44Z","abstract":[{"lang":"eng","text":"Neural machine translation (NMT) is a widely popular text generation task, yet there is a considerable research gap in the development of privacy-preserving NMT models, despite significant data privacy concerns for NMT systems. Differentially private stochastic gradient descent (DP-SGD) is a popular method for training machine learning models with concrete privacy guarantees; however, the implementation specifics of training a model with DP-SGD are not always clarified in existing models, with differing software libraries used and code bases not always being public, leading to reproducibility issues. To tackle this, we introduce DP-NMT, an open-source framework for carrying out research on privacy-preserving NMT with DP-SGD, bringing together numerous models, datasets, and evaluation metrics in one systematic software package. Our goal is to provide a platform for researchers to advance the development of privacy-preserving NMT systems, keeping the specific details of the DP-SGD algorithm transparent and intuitive to implement. We run a set of experiments on datasets from both general and privacy-related domains to demonstrate our framework in use. We make our framework publicly available and welcome feedback from the community."}]}