Amazon-lex returning clarification prompt for all inputs

1
votes

I am trying to create a chatbot in amazon lex but no matter what input I type it returns the clarification prompt to me. I have currently configured the bot with 2 slots mnemonic and company and with 2 sample utterances in a single intent requestParameterInfo

what is the {mnemonic}​ value for {company}​

what is the {mnemonic}​ value for {company}​ for the year {date}

any suggestions why this might be happening ?

1
amazon-lex
i'm having the opposite issue... no matter what I type in, Lex identifies some intent - never brings up a clarification prompt... - Radu Simionescu
@RaduSimionescu make sure slots are required. - AntonIva
@AntonIva -in my case it was actually a LEX bug that got fixed in the meantime - Radu Simionescu

1 Answers

0
votes

Couple things to check:

  • Make sure your Slots have a number of good example values.
  • Make sure you have built the bot and are testing on the current version.
  • Under monitoring, check the missed utterances and add them as sample utterances.
  • Check your Lambda initialization and validation hook is not overriding or otherwise messing with the request.

If none of these are the case, it could be that the values you are providing for {mnemonic} may not be able to be identified by Lex. I've seen this issue when trying to detect yes and no as a slot, and wouldn't be surprised if other words, or non-words are not detected.