When building a chatbot as conversational assistant, we may need some of user’s information in order to answers and give suggestions to the user in a right context.
This proccess of collecting the user’s required information is called slot filling. (If you want to understand better about slots, you may want to check out the previous blog post here)
The situation when users’ required information (required slots) are fulfill in an initial message is called a “happy path”, but in a real-world problems things may get complicated and there are other paths where the user does not directly all the required informations. And this is where we might need FormActions.
FormActions in RASA is an action where it iterates and bring the conversations until all the required slots are filled and also helps reducing the amount of data train in dialogue management.
In FormActions, we can handle certain conditions by adding logic. We can build any kind of logic to suit the purpose of conversational assistant that we want to build.
In RASA, we can try to run our bot using the debug flag to see for ourselves whether the FormActions works great.
Here is a tutorial on how to use FormActions in RASA and a brief explanation of the contextual assistants using RASA Forms: