翻译技术速递The Present and Future of Customized Machine Translation
定制机器翻译的现状与未来
Article by Jourik Ciesielski.
Machine translation finding its way
The year is 2023. Six years after the big neural MT push of 2017, it seems appropriate to say that machine translation (MT) has finally found its way in the localization industry. Most MT providers are producing reasonably acceptable baseline quality and MT solutions have never been more accessible. As a result, MT is becoming a reality in many organizations. What’s more, MT technology has reached a certain level of maturity in terms of customization and training. The barrier to working with legacy data and using it to train custom models and deploy proprietary MT programs — both large and small — is no longer as insurmountable as it once was. The purpose of this article is to provide an update of the state of the current MT technology market and a summary of everything you need to know to deploy trained MT models in your localization processes.
Navigating the technology market
Although Google Translate (public) and SYSTRAN (corporate) once dictated the law of the land of the MT market, things have changed. The technology market has exploded and has enabled every organization to find the right MT solutions partner, even if the quest might be a long and intense one. There are now three primary technology tracks to choose from: cloud services, SaaS providers, and open-source frameworks.
The MT tool landscape. Source: Nimdzi Language Technology Atlas
作者：Jourik Ciesielski
机器翻译找到出路
现在是2023年。在2017年大规模神经MT推动六年之后，似乎可以说机器翻译（MT）终于在本地化行业找到了自己的方式。大多数机器翻译供应商正在生产合理可接受的基线质量，机器翻译解决方案从未如此容易获得。因此，MT在许多组织中正在成为现实。更重要的是，MT技术在定制和培训方面已经达到了一定的成熟度。使用遗留数据并使用它来训练自定义模型和部署专有MT程序（无论大小）的障碍不再像以前那样不可逾越。本文的目的是提供当前MT技术市场的最新状态，并总结在本地化流程中部署经过训练的MT模型所需了解的一切。
驾驭技术市场
虽然Google Translate（公共）和SYSTRAN（企业）曾经主宰了机器翻译市场的法律，但事情已经发生了变化。技术市场已经爆炸式增长，使每个组织都能够找到合适的MT解决方案合作伙伴，即使这可能是一个漫长而激烈的追求。现在有三个主要的技术路线可供选择：云服务、SaaS提供商和开源框架。
MT工具的前景。来源：尼姆兹语言技术图集
以上中文文本为机器翻译，存在不同程度偏差和错误，请理解并参考英文原文阅读。
阅读原文