Online learning for effort reduction in interactive neural machine translation

Autor: Peris-Abril, Álvaro, Casacuberta Nolla, Francisco
Rok vydání: 2019
Předmět:
FOS: Computer and information sciences
Machine translation post-editing
Machine translation
Computer science
Interactive machine translation
media_common.quotation_subject
02 engineering and technology
computer.software_genre
Machine learning
01 natural sciences
Theoretical Computer Science
Domain (software engineering)
Reduction (complexity)
Adaptive system
0103 physical sciences
0202 electrical engineering
electronic engineering
information engineering
Quality (business)
010301 acoustics
Protocol (object-oriented programming)
media_common
Domain adaptation
Computer Science - Computation and Language
Neural machine translation
business.industry
Deep learning
020206 networking & telecommunications
Human-Computer Interaction
Online learning
Artificial intelligence
business
Computation and Language (cs.CL)
LENGUAJES Y SISTEMAS INFORMATICOS
computer
Software
Zdroj: RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
instname
ISSN: 0885-2308
DOI: 10.1016/j.csl.2019.04.001
Popis: [EN] Neural machine translation systems require large amounts of training data and resources. Even with this, the quality of the translations may be insufficient for some users or domains. In such cases, the output of the system must be revised by a human agent. This can be done in a post-editing stage or following an interactive machine translation protocol. We explore the incremental update of neural machine translation systems during the post-editing or interactive translation processes. Such modifications aim to incorporate the new knowledge, from the edited sentences, into the translation system. Updates to the model are performed on-the-fly, as sentences are corrected, via online learning techniques. In addition, we implement a novel interactive, adaptive system, able to react to single-character interactions. This system greatly reduces the human effort required for obtaining high-quality translations. In order to stress our proposals, we conduct exhaustive experiments varying the amount and type of data available for training. Results show that online learning effectively achieves the objective of reducing the human effort required during the post-editing or the interactive machine translation stages. Moreover, these adaptive systems also perform well in scenarios with scarce resources. We show that a neural machine translation system can be rapidly adapted to a specific domain, exclusively by means of online learning techniques.
The authors wish to thank the anonymous reviewers for their valuable criticisms and suggestions. The research leading to these results has received funding from the Generalitat Valenciana under grant PROMETEOII/2014/030 and from TIN2015-70924-C2-1-R. We also acknowledge NVIDIA Corporation for the donation of GPUs used in this work.
Databáze: OpenAIRE
Externí odkaz: