Common Transformer-based translation models use embedding, which can capture the semantic relationships of words in high-dimensional space, and the extracted features are directly transmitted to the attention mechanism of multiple heads. Aiming at the issue of how to better extract the correlation between English and French languages and improve translation quality in English French translation, an English French translation model based on Transform is designed. Combined with the doorstep loop unit, it has the ability to extract semantic information from consecutive sequences, thereby better extracting language information. In terms of improving translation performance, using mask tensors can prevent future information from being used in translation in advance and using greedy algorithms to generate sequences, improve training efficiency and focus translation attention on important information. In terms of optimizing models, using Logsoftmax can alleviate the overflow or underflow problems of Softmax, Replace traditional Transformer's Softmax. and introduce a teacher_ Forcing to avoid errors in the sequence generation process and improve the accuracy of French translation. The experimental results show that under the wmt2014 English French dataset and self-collected dataset, relatively good results have been achieved in effectively extracting association features between two languages and improving translation tasks, with a BLEU of about 0.804.

Xu Bowei. Design of Transformer Based English French Translation Model. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 5: 30-38. https://doi.org/10.25236/AJCIS.2023.060505.

[1] Ren Wen. The Challenge and Orientation of Machine Translation Ethics [J]. Shanghai Translation, Issue, 2019(5):46-52+95.

[2] Zheng Man, Hu Xianyao. The Theory, Development, and Prospect of Constructive Translation Theory [J]. Journal of Xi'an International Studies University, 2022, 30(3):5.

[3] Fan Wuqiu, Wang Yu. Spiritual Interaction between Translator and Text: A Bottleneck to Be Breakthrough in Machine Translation [J]. Foreign Language Teaching Theory and Practice, 2022(03): 128-137.

[4] Liu Qun. Survey on statistical machine translation [J]. Journal of Chinese Information Processing, 2003, 17(4):1-12.

[5] Gao Minghu, Yu Zhiqiang. Asummary review of neural machine translation [J]. Journal of Yunnan University of Nationalities (Natural Sciences Edition), 2019, 28(1):72-76. 

[6] Kalchbrenner N, Blunsom P. Recurrent continuous translation models [C] //Proc of 2013 Conference on Empirical Methods in Natural Language Processing, 2013:1700-1709.

[7] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need [C]// Proc of the 31st International Conference on Neural Information Processing Systems, 2017:5998-6008.

[8] Artetxe M, Labaka G, Agirre E, et al.Unspervised neural machine translation [J]. arXiv: 1710. 11041v2, 2018.

[9] Lamole G, Ott M, Conneau A, et al. Phrase-based & neural unsupervised machine translation [C] // Proc of the 2018 Conference on Empirical Methods in Natural Language Processing, 2018:5039-5049.

[10] Ren S, Wu Y, Liu S, et al. A Retrieve-and-Rewrite initialization method for unsupervised machine translation[C]//Proc of the 58th Annual Meeting of Association for Computational Linguistics, 2020:3498-3504.

[11] Wang Tao, Xiong Deyi. Enhancing Neural Machine Translation with Predefined Bilingual Pairs [J]. Chinese Journal of Information Technology, 2022, 36 (06): 36-43.

[12] Tao C., Gao S., Shang M., Wu W, Zhao D, Yan R, Get the point of my utterance! Learning towards effective responses with multi-head attrntion mechanism [C/OL]. [2022-06-12]. http: // doi. org/ 10. 24963/ ijcai.2018/614

[13] Wang Xu, Jia Hao, Ji Baijun, Duan Xiangyu. Neural Machine Translation Based on Dictionary Model Fusion [J]. Computer Engineering and Science, 2022, 44 (08): 1481-1487

[14] Feng Zhiwei. Three Methods of Generating Word Vector [J]. Foreign Language Audiovisual Teaching, 2021 (01): 18-26+3.

[15] Gengchen Mai, Krzysztof Janowicz, Yingjie Hu, Song Gao, Bo Yan, Rui Zhu, Ling Cai & Ni Lao. A review of location encoding for GeoAI: methods and applications [J]. International Journal of Geographical Information Science, 2022, 36(4):639-673.

[16] Jiang Menghan, Li Shaomei, Zheng Honghao, Zhang Jianpeng. A Rumor Detection Model Based on Improved Location Coding [J]. Computer Science, 2022, 49 (08): 330-335.

[17] Huang Hao, Zhou Lihua, Huang Yaqun, Jiang Yiting (2022). Early detection of false information based on hybrid depth model Journal of Shandong University (Engineering Edition), 2022 (52 (04)), 89 – 98+109.

[18] Kang Xiaomian, Zong Chengqing. Neural Machine Translation Based on Textual Structure Multitask Learning [J]. Journal of Software, 2022, 33(10):3806-3818.

[19] Liu Hao, Hong Yu, Zhu Qiaoming. Unsupervised Domain Adaptive Machine Reading Comprehension Method [J]. Journal of Computer Science, 2022, 45(10):2133-2150.