Abstract
For many years, machine translation has been one of the most important and challenging topics in the field of natural language processing. In this work, we discuss and implement the real-time synchronous translation method, and focus on the key technologies to be solved in the translation generation of real-time synchronous translation method. The optimal template selection and phrase translation are the key factors affecting template machine translation. We improve the selection of the optimal template by using the methods of text template direct matching and template selection. In addition, the sequence-to-sequence model based on Recurrent Neural Network (RNN) has achieved good results in the task of translation text generation, but most of these models have the problems of text repetition, and exposure deviation. Aiming at the repetition problems, we propose a hybrid attention composed of stored attention and decoded self-attention, which is overcome by storing historical attention. In order to solve the problem of exposure bias and correct the loss function, we design a new training method based on reinforcement learning. In the experiment, we test the model on the ChinaDaily dataset and take Recall-Oriented Understudy for Gisting Evaluation (ROUGE) as the evaluation index. The results show that mixed attention can greatly improve the repetition problem, the exposure deviation can be eliminated with reinforcement learning, and the integrated model surpasses the State-of-the-Art algorithms in the test set.
Similar content being viewed by others
Data availability
The data used to support the findings of this study is available from the corresponding author upon the reasonable request.
References
Li, S., Zhang, S., Jia, C., et al. (2020). Direct speech-to-image translation. IEEE Journal of Selected Topics in Signal Processing, 14, 517–529. https://doi.org/10.1109/JSTSP.2020.2987417.
Chen, J., Wu, Y., Jia, C., et al. (2020). Customizable text generation via conditional text generative adversarial network. Neurocomputing, 416, 125–135. https://doi.org/10.1016/j.neucom.2018.12.092.
Stoll, S., Camgoz, N. C., Hadfield, S., et al. (2020). Text2Sign: Towards sign language production using neural machine translation and generative adversarial networks. International Journal of Computer Vision, 128, 891–908. https://doi.org/10.1007/s11263-019-01281-2.
Casado, L. F., García-Gutiérrez, J. V., Massagué, I., et al. (2015). Switching to second-generation tyrosine kinase inhibitor improves the response and outcome of frontline imatinib-treated patients with chronic myeloid leukemia with more than 10% of BCR-ABL/ABL ratio at 3months. Cancer Medicine, 4, 995–1002. https://doi.org/10.1002/cam4.440.
Och, F., & Ney, H. (2004). The alignment template approach to statistical machine translation. Computational Linguistics, 30, 417–449. https://doi.org/10.1007/comlig.2004-30417.
Daybelge, T., & Cicekli, I. (2011). A ranking method for example-based machine translation results by learning from user feedback. Applied Intelligence, 35, 296–321. https://doi.org/10.1007/s10489-010-0222-7.
Wang, J. D., Chen, Z. X., & Huang, H. Y. (2001). Intelligent case based machine translation system, computational linguistics and intelligent text processing. IEEE International Conference on Systems Structure, 25, 18–24. https://doi.org/10.1007/3-540-44686-9_21.
Liu, Y., & Guo, S. (2016). Generation and dynamics analysis of N-scrolls existence in new translation-type chaotic systems. Chaos, 26, 113–114. https://doi.org/10.2991/msam-18.2018.50.
Zhang, J., Chen, D., Han, G., et al. (2020). SSNet: Structure-Semantic Net for Chinese typography generation based on image translation. Neurocomputing, 371, 15–26. https://doi.org/10.1016/j.neucom.2019.08.072.
Liu, M., Li, L., Hu, H., et al. (2020). Image caption generation with dual attention mechanism. Information Processing & Management, 57, 102178–102191. https://doi.org/10.1016/j.ipm.2019.102178.
Zhao, X., Shang, P., & Huang, J. (2016). Mutual-information matrix analysis for nonlinear interactions of multivariate time series. Nonlinear Dynamics, 88, 1–11. https://doi.org/10.1007/s11071-016-3254-7.
Brown, S. E., & Dusing, S. C. (2019). Knowledge translation lecture: Providing best practice in neonatal intensive care and follow-up. Pediatric Physical Therapy, 31, 308–314. https://doi.org/10.1097/PEP.0000000000000634.
Mcbride, C. M., Abrams, L. R., & Koehly, L. M. (2015). Using a historical lens to envision the next generation of genomic translation research. Public Health Genomics, 18, 272–282. https://doi.org/10.1159/000435832.
Yang, G., & Yu, T. (2018). Generation of isogenic single and multiplex gene knockout mice by base editing-induced STOP. Science Bulletin, 63, 19–25. https://doi.org/10.1016/j.scib.2018.07.002.
Huo, L. Z., Boschetti, L., Sparks, A. M., et al. (2017). Deforestation and industrial forest patterns in Colombia: A case study. AGU Fall Meeting Abstracts, 12, 1632–1637. https://doi.org/10.1189/0004303.
Cho, H., Yu, L., & Abdi, S. (2013). Automatic generation of transducer models for bus-based MPSoC design. IEEE Transactions on Computers, 62, 211–224. https://doi.org/10.1109/TC.2012.157.
Zhao, Y., Zeng, D., Socinski, M. A., et al. (2015). Reinforcement learning strategies for clinical trials in nonsmall cell lung cancer. Biometrics, 67, 1422–1433. https://doi.org/10.1111/j.1541-0420.2011.01572.x.
Ardi, T., Tambet, M., Dorian, K., et al. (2017). Multiagent cooperation and competition with deep reinforcement learning. PLoS ONE, 12, 1–12. https://doi.org/10.1371/journal.pone.0172395.
Olivecrona, M., Blaschke, T., Engkvist, O., et al. (2017). Molecular de-novo design through deep reinforcement learning. Journal of Cheminformatics, 9, 1–14. https://doi.org/10.1186/s13321-017-0235-x.
Gershman, S. J., & Daw, N. D. (2017). Reinforcement learning and episodic memory in humans and animals: An integrative framework. Annual Review of Psychology, 68, 101–128. https://doi.org/10.1146/annurev-psych-122414-033625.
Xue, B. P., Berseth, G., & Panne, M. (2016). Terrain-adaptive locomotion skills using deep reinforcement learning. ACM Transactions on Graphics, 4, 1–12. https://doi.org/10.1145/2897824.2925881.
See, A., Liu, P. J., & Manning, C. D. (2017). Get to the point: Summarization with pointer-generator networks. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Vol. 1, pp. 1073–1083). https://doi.org/10.18653/v1/P17-1099
Peng, X. B., Abbeel, P., Levine, S., et al. (2018). DeepMimic: Example-guided deep reinforcement learning of physics-based character skills. ACM Transactions on Graphics, 37, 1–14. https://doi.org/10.1145/3197517.3201311.
Conde, M., Chisini, L. A., Demarco, F. F., et al. (2016). Stem cell-based pulp tissue engineering: Variables enrolled in translation from the bench to the bedside, a systematic review of literature. International Endodontic Journal, 49, 543–550. https://doi.org/10.1111/iej.12489.
Michael, L. (2015). Reinforcement learning improves behaviour from evaluative feedback. Nature, 521, 445–451. https://doi.org/10.1038/nature14540.
Duan, G., et al. (2021). Improving neural machine translation model with deep encoding information. Cognitive Computation, 3, 1–9. https://doi.org/10.1007/s12559-021-09860-7.
Wang, F., et al. (2019). Hybrid attention for Chinese character-level neural machine translation. Neurocomputing, 358, 44–52. https://doi.org/10.1016/j.neucom.2019.05.032.
Moschoglou, S., Ploumpis, S., Nicolaou, M. A., et al. (2020). 3DFaceGAN: Adversarial nets for 3D face representation, generation, and translation. International Journal of Computer Vision, 128, 2534–2551. https://doi.org/10.1007/s11263-020-01329-8.
Moreira, R. B., Peixoto, R. D., Sousa, T., et al. (2015). Next-generation sequencing (NGS) in metastatic breast cancer (mBC) patients: Translation from sequence data into clinical practice. Journal of Clinical Oncology, 11, 13–60. https://doi.org/10.1200/jco.2015.33.28_suppl.133.
Yang, Y., Dan, X., Qiu, X., et al. (2020). FGGAN: Feature-guiding generative adversarial networks for text generation. IEEE Access, 8, 105217–105225. https://doi.org/10.1109/ACCESS.2020.2993928.
Xie, J., et al. (2020). Chinese text classification based on attention mechanism and feature-enhanced fusion neural network. Computing, 102, 1–1. https://doi.org/10.1007/s00607-019-00766-9.
Xiao, Y., et al. (2022). Hybrid attention-based transformer block model for distant supervision relation extraction. Neurocomputing, 470, 29–39. https://doi.org/10.1016/j.neucom.2021.10.037.
Kim, T., Yun, Y., & Kim, N. (2021). Deep learning-based knowledge graph generation for COVID-19. Sustainability, 13, 2276–2285. https://doi.org/10.3390/su13042276.
Hu, T., & Meinel, C. (2021). Masked hard coverage mechanism on pointer-generator network for natural language generation. In 13th International Conference on Agents and Artificial Intelligence (Vol. 13, pp. 1–12). https://doi.org/10.5220/0010341211771183
Stoll, S., Camgoz, N. C., Hadfield, S., et al. (2020). Text2Sign: Towards sign language production using neural machine translation and generative adversarial networks. International Journal of Computer Vision, 128, 891–908. https://doi.org/10.1007/s11263-019-01281-2.
Paul, S., & Saha, S. (2020). CyberBERT: BERT for cyberbullying identification. Multimedia Systems, 2020, 1–8. https://doi.org/10.1007/s00530-020-00710-4.
Belogolovsky, S., et al. (2021). Inverse reinforcement learning in contextual MDPs. Machine Learning, 3, 1–1. https://doi.org/10.1007/s10994-021-05984-x.
Jan, H., & Jensen. (2019). A graph-based genetic algorithm and generative model/Monte Carlo tree search for the exploration of chemical space. Chemical Science, 1, 1–1. https://doi.org/10.1039/c8sc05372c.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflicts of interest.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Ke, X. English synchronous real-time translation method based on reinforcement learning. Wireless Netw (2022). https://doi.org/10.1007/s11276-022-02910-4
Accepted:
Published:
DOI: https://doi.org/10.1007/s11276-022-02910-4