Azure translation issue(from English to Chinese), unnecessary extra spaces appear in the generated document when a glossaries is used.

Luoye An 5

We use the Azure Translator tool for document translation (translating English documents into Chinese documents). When a glossary is used, the generated Chinese document contains unnecessary spaces between each Chinese character. If the glossary is not used, the translated Chinese document appears normal.

Refer to the link below: https://zcusa.951200.xyz/en-au/azure/ai-services/translator/document-translation/how-to-guides/create-use-glossaries

santoshkc 11,535 Reputation points Microsoft Vendor

2024-12-02T11:03:59.4966667+00:00

Hi @Luoye An,

Thank you for reaching out to Microsoft Q&A forum and highlighting the issue with unnecessary spaces appearing in the translated document when using glossaries in Azure Translator. This behaviour could be due to how the glossary terms are being applied during the translation process, especially for languages like Chinese that do not use spaces between characters naturally. To address this, ensure that your glossary file uses the correct format (e.g., tab-separated values) and check if the terms in the glossary are appropriately aligned with the source text. Additionally, verify that the translation request specifies the glossary accurately in the JSON payload, as mismatched configurations can lead to formatting inconsistencies.

I hope you understand! Thank you.
Luoye An 5 Reputation points

2024-12-03T06:52:34.82+00:00

We don’t have unnecessary spaces in our glossary. Here is a portion of the TSV glossary; can you help check if there are any issues? @santoshkc

3D geometry mode 3D几何模型

non-watertight geometry 不完全封闭的脏几何

pre-processing 前处理

external flow simulations 外流仿真

conformal volume meshing methods Conformal体积网格划分方法

toggled 勾选

outer flow boundary 外部流场域的边界

engineering simulation 工程仿真/技术

Mosaic-enabled 基于Mosaic

geometry options 几何结构参数

heat transfer 换热或传热

tube air exchangers. 管式换热器

heat exchanger 换热器

heat exchanger system. 热交换器系统

water-intensive 高耗水

heat exchanger tubes 热交换器管
santoshkc 11,535 Reputation points Microsoft Vendor

2024-12-05T14:21:25.06+00:00

Hi @Luoye An,

Thank you for your follow-up query.

The TSV glossary you provided appears to be formatted correctly. However, the issue of unnecessary spaces might be related to how Azure Translator handles Chinese text, as Chinese characters do not typically use spaces. Even though the glossary seems fine, check if there are any hidden characters or extra spaces in the glossary file itself. Also, ensure that the file is saved in UTF-8 encoding. If the issue continues, try testing with a smaller glossary or specific terms to isolate the cause. Still if you face the issues, I request you to raise a support request through the Azure portal to help further diagnose the problem.

Thank you.

Share via

Azure translation issue(from English to Chinese), unnecessary extra spaces appear in the generated document when a glossaries is used.

Your answer