How to detect fallback language in NLP of email bot?
We would like to use the NLP to automatically detect the language of the emails received from our customers in the our Customer Service (CS) application.
In particular we would like to distinguish the cases when the algorithm uses the fallback language (e.g. English ) from the cases where it detects the same language (english).
Our idea was to disable the fallback language and distinguishing the situations based on the output result of NLP: empty (when it was not able to detect the language) and non-empty (when it detects the language), but in the CS we are not able to disable this option (every time that we enable it, then after the refresh of the rule the checkbox appears checked again).
Can you help with this?
***Edited by Moderator Marije to add bug BUG-884518 Details***