Question
Siemens Technology and Services Pvt. Ltd.
IN
Last activity: 29 Nov 2023 9:41 EST
NLP Language detection
Hello,
We are maintaining Models for 2 different languages. Sometimes the detected language is different than the maintained Model language. Hence the entitites are not detected.
Is there a way to auto detect the language as well as default the language to EN incase the detected language is different than these 2.
Also, how to use the Language detected by publisher feature, I am looking for examples here.
Accepted Solution
Updated: 29 Nov 2023 9:41 EST
Pegasystems Inc.
GB
To auto-detect the language and set a fallback language to English, follow these steps:
- In the Records panel, click Decision > Text Analyzer.
- Click the Advanced tab.
- In the Language settings section, select Automatically detect language.
- Select Enable fallback language if language undetected and specify English as the fallback language.
To use the "Language detected by publisher" feature:
- In the Language settings section of the Text Analyzer's Advanced tab, select the Language detected by publisher radio button.
- This option uses the language metadata tag (lang:) of the incoming records for language detection.
By following these steps, your Text Analyzer will automatically detect the language and use English as a fallback language if the detected language is different from the two maintained languages. It will also use the language metadata tag provided by the publisher for language detection.