Hi,
I am using the carousel template to display a list of items. We have two buttons for each item and one of them is ‘Know More’. Clicking ‘Know more’ will trigger a message node that displays more details for the item. After this, the dialog task will reach the end of the dialog.
My question is, how can we handle a scenario when the user clicks on the ‘Know more’ button again for a different item or for the same item?
Thanks.
Christian,
There are three strategies.
-
Depending on the chat client, then you may be able to modify the UI after the user has clicked the button to either make it disappear or be disabled.
-
All the buttons do is simulate the user typing text, so the same “issue” exists even if the user explicitly types “know more”. So you really need to consider this case regardless of whether there are buttons or not. That means either having explicitly primary intent training for that phrase, or you don’t and let the utterance roll through to the default dialog where you could check - perhaps through a trait or a hidden list of items entity.
-
The payload of a button can be different from the text displayed on the button, which means that more text can be sent to the NLP engine. So each button could have a payload that contains context, e.g. “know more about changing your password”, and that extra “context” can be picked up by primary intent identification when you are outside of the original dialog.