I believe one of the reasons for using “Continuation of” is to induce a speaker's possible utterances. However, I think it would be better if we are able able to link the same utterance to different actions in each specific views.
1. In the concept "Order" View, it has goal "AddressBook1", route action "FillOutHomeAddress"
[g: AddressBook1: continue: Order, r: FillOutHomeAddress] Fill the address
on the other hand,
2. In the concept "Receipt" View, it has goal "AddressBook2", route action "FillOutMostRecentAddress"
[g: AddressBook2: continue: Receipt, r: FillOutMostRecentAddress] Fill the address
I'm not sure why there's a reason to use "continuation of" if the same utterance crashes when it comes out in different situations.
what's the difference from just training without a continuation of?
Please sign in to leave a comment.