什么是 Translator 服务?What is the Translator service?

Translator 是一种基于云的机器翻译服务,是用于构建智能应用的 Azure 认知服务系列认知 API 的一部分。Translator is a cloud-based machine translation service and is part of the Azure Cognitive Services family of cognitive APIs used to build intelligent apps. “翻译”可以轻松地集成到应用程序、网站、工具和解决方案中。Translator is easy to integrate in your applications, websites, tools, and solutions. 通过它,你可以添加超过 70 种语言的多语言用户体验,并且可以在具有任何操作系统的任何硬件平台上将其应用于文本翻译。It allows you to add multi-language user experiences in more than 70 languages, and can be used on any hardware platform with any operating system for text translation.

关于 Microsoft TranslatorAbout Microsoft Translator

“翻译”为各种 Microsoft 产品和服务提供支持,并已在全球数千家企业的应用程序和工作流中使用,使其内容可传播到全球的受众。Translator powers a number of Microsoft products and services, and is used by thousands of businesses worldwide in their applications and workflows, which allows their content to reach a global audience.

由 Translator 提供支持的语音翻译也可通过 Azure 语音服务获取。Speech translation, powered by Translator, is also available through the Azure Speech service. 它将语音翻译 API 和自定义语音服务的功能组合成了一项统一的可完全自定义的服务。It combines functionality from the Translator Speech API and the Custom Speech Service into a unified and fully customizable service.

语言支持Language support

Translator 为文本翻译、音译、语言检测和字典提供多语言支持。Translator provides multi-language support for text translation, transliteration, language detection, and dictionaries. 请参阅语言支持以获取完整的列表,或者通过 REST API 以编程方式访问列表。See language support for a complete list, or access the list programmatically with the REST API.

Microsoft Translator 神经机器翻译Microsoft Translator Neural Machine Translation

神经机器翻译 (NMT) 是采用 AI 的高质量机器翻译的新标准。Neural Machine Translation (NMT) is the new standard for high-quality AI-powered machine translations. 它代替的是旧式统计机器翻译 (SMT) 技术,该技术在 2010-2020 中期的几年中达到了质量顶峰。It replaces the legacy Statistical Machine Translation (SMT) technology that reached a quality plateau in the mid-2010s.

与 SMT 相比,NMT 不仅能够从原始翻译质量评分的立场提供更好的翻译,而且听起来更流畅、更类似于人类。NMT provides better translations than SMT not only from a raw translation quality scoring standpoint but also because they will sound more fluent and human. 这种流畅性的主要原因在于 NMT 使用一个句子的完整语境来翻译单词。The key reason for this fluidity is that NMT uses the full context of a sentence to translate words. SMT 仅考虑每个单词前面和后面几个单词的直接语境。SMT only took the immediate context of a few words before and after each word.

NMT 模型是该 API 的核心,对最终用户不可见。NMT models are at the core of the API and are not visible to end users. 唯一明显的区别是改进的翻译质量,尤其是针对中文、日语和阿拉伯语等语言。The only noticeable difference is improved translation quality, especially for languages such as Chinese, Japanese, and Arabic.

详细了解 NMT 的工作原理Learn more about how NMT works

后续步骤Next steps