Open w1nk opened 5 years ago
That functionality does not yet exist.
If it's a useful feature for the project, I don't mind assisting in bringing such functionality into existence. I've already done quite a bit of patching so my pytext webservice resembles the luis api close enough during the trigger selection to work, presumably the work is roughly similar to extend the matching inside the script (ie: scan the triggers for type intent and lob those at the service for answer). If this is something you'd see as interesting, can you give me a little bit of guidance around:
1) Is my understanding of the mechanism inside the scripts at all close? ie: does that code resemble the trigger resolution mechanisms as well? 2) Are there standards guidelines I can follow to not get lost in naming convention during code reviews, etc?
If it's not interesting, that'd be super useful to know as well :)
Hello!
I've created a model with a few intents and have it connected to botkit-cms, and then my bot connected to that. So far everything seems to function as expected. When configuring scripts and their associated triggers, I can select the intents from my model and things work great.
Once inside a script, is there also a way to have the conditional actions use my model's intents to choose what to do? So instead of exact/regex matching, use the model intent prediction like the trigger mechanism?
Thanks!