Closed remster85 closed 6 years ago
Overall confidence is based on the amount of the utterance that is related to the matched intent. Entity matches from regular expressions are weighted lower than matches from vocabulary. You can see where this confidence rating is applied in EntityTagger.
In the above utterance, only "weather" and "tokyo" are used as part of the intent, and as such it has a low weight.
thank you for the explanation!
Hi guys, I like the framework a lot and plan to use it and contribute to it. I was playing with the regex example. and executing the command, it returns a low confidence.