What kind of algorithm has been followed by NLU for selecting intent?

I have two different intent. both of them has exact same utterances but only one of them is getting find in ‘event.nlu.intents’ array with confidance 1. but i want to leave this both options to user that what exactly he/she want to do with that.

and also appreciate if someone give me idea how confidence get eventuated.

Hi Nisargkdesai,

Thanks for taking the time of asking a question.

I’m not sure there is an easy solution for this. I believe there is something coming up in that sense, however i couldn’t tell you more i’m not technical myself.

If this an issue you are seeing just a couple times, The way i have seen conversational designers deal with this issue in the past, is they create a “larger intent” that encompasses both “intents” and have them flow through choice skill to choose to validate the true end user intent.

Hope this makes sense and in the meantime, happy bot building
JB

“Larger Intent” that encompasses both “Intents” i did not understand. I did not found that how to create an Intent Hierarchy?

By Choice skill we could separate intent but the thing is that conflict between utterance is something that generates the choices. there are bunch of intent have this kind of conflicts.
I can handle this by create different flow for all of this utterance than Choices and through choices specific intent. but than i need to create total 16 to 18 different flow for all of the utterances.
like if utterance pay load is specific then this intent selection choices appears and then specific intent.

still wondering if i could handle this dynamically then it will be much better and flexible and efficient as well.