wit/sentiment not training

[stergosjk]

Do you want to request a feature, report a bug, or ask a question about wit? Question or Bug

What is the current behavior? I am using both the Understanding and postman to test the term: my manager is a wanka

I have attempted to train it and can see in Utterances that is labelled correctly. ie intent = comment analysis, custom entity 'profanity' appears and also Trait wit/sentiment is negative. However if I go back to Understanding or use postman to test this I only receive the intent and entity correctly. It fails to return the sentiment. It returns no sentiment at all.

If the current behavior is a bug, please provide the steps to reproduce and if possible a minimal demo of the problem. Use postman: https://api.wit.ai/message?v=&q=my manager is a wanka

What is the expected behavior? Expected behaviour is to return intent = comment analsysis, entity profanity with a confidence above 90% and sentiment of negative

If applicable, what is the App ID where you are experiencing this issue? If you do not provide this, we cannot help. 1958910467456264

[patapizza]

Hi @stergosjk, thanks for the note.

At this point, wit/sentiment doesn't get iterative improvements. I'll keep this open for tracking.