Score sentiment - text analytics - Hebrew support issue

[OriBy]

i'm trying to use Score sentiment in PQ , but , when using text in hebrew it returns null , eventough it allegedly supported in the docoumention under the "he" code

-  https://learn.microsoft.com/en-us/azure/ai-services/language-service/sentiment-opinion-mining/langua...

[v-sshirivolu]

Hi @OriBy ,
Thank you for reaching out to Microsoft Fabric Community Forum.
Steps to try:

Language Parameter:
Make sure you are specifically setting he (the ISO 639-1 language code for Hebrew) as the language parameter in the ScoreSentiment function. This avoids relying on automatic language detection, which may not correctly identify Hebrew.

Encoding Issue:
If your source text includes special Hebrew characters, confirm the encoding is correct. Consider cleaning or re-importing the data using UTF-8 encoding.

Service Region + Key:
For a custom Cognitive Service, check that:

The region matches your resource (e.g., westeurope, eastus).

Your API key is valid and authorized for sentiment analysis.

Length or Format Issues:
Very short, ambiguous, or incorrectly tokenized sentences may return null. Try using:

Longer Hebrew text samples

Correct punctuation

Test with REST API Directly:
To determine if the problem is with Power BI or the service itself:

Send a sample Hebrew sentence using Language Studio or Postman via the REST API.

If the API provides a sentiment score, the issue is likely on the Power BI side.

Thank you.

[v-sshirivolu]

Hi @OriBy ,

I hope the information provided above assists you in resolving the issue. If you have any additional questions or concerns, please do not hesitate to contact us. We are here to support you and will be happy to help with any further assistance you may need.

[v-sshirivolu]

Hi @OriBy ,

I hope the above details help you fix the issue. If you still have any questions or need more help, feel free to reach out. We’re always here to support you

[v-sshirivolu]

HI @OriBy ,
I wanted to check if you had the opportunity to review the information provided. Please feel free to contact us if you have any further questions.

[Nasif_Azam]

Hey @OriBy ,

Few things to consider:

1. Check Azure Cognitive Services Settings: Ensure that the Azure Cognitive Services endpoint you're using has sentiment analysis support for Hebrew enabled. While the "he" language code is mentioned in the documentation, there could be specific configuration issues or service limitations affecting its functionality.

2. Text Encoding and Cleanliness: Verify that the Hebrew text is correctly encoded. Sometimes, special characters or encoding mismatches can lead to unexpected results in text analytics.

3. Sentiment Analysis API: You can also try using a different approach, such as calling the sentiment analysis API directly through Power Automate if Power Query's native function isn’t working as expected.

4. Limitations in Language Detection: Sometimes, the language detection might not be accurate for Hebrew, leading to null values. Try manually specifying the language code or preprocessing the text to ensure it's correctly identified as Hebrew.

If you found this solution helpful, please consider accepting it and giving it a kudos (Like) it’s greatly appreciated and helps others find the solution more easily.

Best Regards,
Nasif Azam