Intent classification with wrong intent

I’ve been training my rasa model with few questions, but actually there is misclassification of intents which are not required.

For example I have two questions: Q1. What is your biological gender? Q2. What is your identifying gender?

Now the problem here is, the training examples for both are same, eg. Male, Female, Prefer not to say, etc.

So when the user types the “identifying gender” , the “biological gender” slot gets filled, that is not intended.

Can Anyone tell me how to resolve this, I’ve tried with Form actions and also changes like not_intent and ignore_intent, but none of them seems to work.

did I understand your problem correctly? you are writing 2 similar things and want them to be grouped in 2 different ways?

solution for this is not a roles&groups?

Roles and Groups

“I want a flight from [Berlin]{“entity”: “location”, “role”: “origin”} to [SF]{“value”: “San Francisco”, “entity”: “location”, “role”: “destination”}”

in there there are 2 cities grouped in 2 different ways

let us take the following example I have examples belonging to 2 separate intents like this

The DietClassifier is misclassifying the intents. How can we avoid this confusion?

- intent: dont_understand
  examples: |
    - no lo he entendido (i don't uderstand it)
    - no lo entiendo 
    - no me aclaro
    - no entiendo nada (i don't understand anything)
    - no entiendo que has dicho
    - no, no lo entiendo
    - no entiendo muchas cosas 
    - no he entendido nada de lo que has dicho
    - nada de lo que has dicho tiene sentido
    - no tiene sentido (it doesn't make sense)
    - ... +20
- intent: deny
  examples: |
    - No
    - no
    - no lo creo (i don't think so)
    - lo dudo mucho
    - Nop
    - Nopp
    - ni pensarlo
    - no por favor (no please)
    - no para nada ```

well I do not understand that language, but maybe it will help? Training Data & Rules – Rasa Learning Center

please take a look on video at 17:23, if that will not help, maybe we need to wait for someone smarter

there is also second case: Language Support (rasa.com)

“Even more so when training word embeddings from scratch, more training data will lead to a better model! If you find your model is having trouble discerning your inputs, try training with more example sentences.”