Hello,
Please see below query. I was searching for some query using try model and I got the result as below screenshot,
It is pointing to update the catalog item wordings intent with 100% threshold.
If I ask the same question in Virtual Agent, It is pointing to different intent and asking to confirm,
The question should not hit this intent(Any query that contains the word RITM should not hit Search Service Catalogs intent). I didn't understand why the results vary using try model and virtual agent when everything is in published state.
Could you please help on this and also "Any query that contains the word RITM should not hit Search Service Catalogs intent" - I tried to write negative vocabulary like this - ^(?!.*\bRITM\w+\b). But nothing worked.
Please answer these two qns.
Thanks,
Satish
I would assume in your utterances of that topic you have similar to this:
Or even for an Incident which means you have an associated entity attached to it like these as I am assuming you cloned the OOB Topic
If so, as you can see below that besides the utterances that is configured in other locations:
As to question two, Solved: block words a user may ask from specific intents - ServiceNow Community but I think if you look at the entities you wont need to do that.
