官方技术资料：Microsoft（第 20 页）
官方技术资料 · Microsoft：共 312 条，本页为第 286–300 条。
Precise answering
Understand Precise answering feature available in custom question answering.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：agent · agent-frameworks
链接复核时间：2026-09-24
Custom question answering project lifecycle
Custom question answering learns best in an iterative cycle of model changes, utterance examples, deployment, and gathering data from endpoint queries.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：agent · agent-frameworks
链接复核时间：2026-09-24
Assertion detection
Learn about assertion detection.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：supervised-learning · privacy
链接复核时间：2026-09-24
Utilizing Fast Healthcare Interoperability Resources (FHIR) structuring in Text Analytics for Health
Learn about Fast Healthcare Interoperability Resources (FHIR) structuring
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：supervised-learning · privacy
链接复核时间：2026-09-24
Supported Text Analytics for health entity categories
Learn about categories recognized by Text Analytics for health
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：supervised-learning · privacy
链接复核时间：2026-09-24
Relation extraction
Learn about relation extraction
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：supervised-learning · privacy
链接复核时间：2026-09-24
Audio concepts in Azure Speech in Foundry Tools
An overview of audio concepts in Azure Speech in Foundry Tools.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：speech · multimodal
链接复核时间：2026-09-24
What is a BLEU score?
BLEU is a measurement of the differences between machine translation and human-created reference translations of the same source sentence.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
Custom Translator text translations
Use the Microsoft Translator Hub to build your own machine translation system using your preferred terminology and style.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
Custom Translator Data filtering
When you submit documents to be used for training a custom system, the documents undergo a series of processing and filtering steps.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
What is a Translator custom dictionary?
How to create an aligned document that specifies a list of phrases or sentences (and their translations) that you always want Microsoft Translator to translate the same way. Dictionaries are sometimes also called glossaries or term bases.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
Custom Translator training and modeling?
A model is the system, which provides translation for a specific language pair. The outcome of a successful training is a model. To train a model, three mutually exclusive data sets are required training dataset, tuning dataset, and testing dataset.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
What are Custom Translator parallel documents?
Parallel documents are paired documents where one is the translation of the other. One document in the pair contains sentences in the source language and the other document contains these sentences translated into the target language.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
Sentence pairing and alignment in parallel documents
During the training execution, sentences present in parallel documents are paired or aligned. Custom Translator learns translations one sentence at a time, by reading a sentence and translating it. Then it aligns words and phrases in these two sentences to each other.
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：transformer · supervised-learning
链接复核时间：2026-09-24
NotesOnThreading.md
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator - onnxruntime/docs/NotesOnThreading.md at main · microsoft/onnxruntime
证据用途：用于理解、实现、选型、可靠性、安全或迁移判断。
使用边界：只直接证明 Microsoft 的实现与声明；预览功能和产品生命周期使用前应复核
关联概念：deployment · inference-optimization
链接复核时间：2026-09-24