This project is a Rasa-based chatbot that collects user preferences and provides personalized recommendations. The chatbot is equipped with various functionalities like handling user interactions, validating form inputs, and more.
- Install Dependencies
- Files Description
- Usage
- Dependencies
pip install -r requirements.txt
- NLU and Training Data:
Defines intents, entities, lookup tables, and training examples for Rasa's Natural Language Understanding.
- Rules and Stories:
Contains the rules and stories which guide the conversation flow of the chatbot.
- Actions:
Implements custom actions used in the conversation, such as validating form inputs and storing user data.
- User Interaction Script:
A script to interact with Furhat
The commands must be used in the root of your Rasa project
- Train the Rasa Model:
rasa train
Be carful if you train a new model you should also change it in the furhat_agent.py program.
Note that a model is already present in the project system.
- Enabling the HTTP API
rasa run --enable-api
WARNING: you might have to re-execute this command if the virtual agent doesn't reply anymore. It mostly occure when you execute for a second time the main program (furhat_agent.py).
Open another terminal and activate the virtual environment to proceed to the next step.
- Run the Rasa Actions Server:
rasa run actions
Open another terminal and activate the virtual environment to proceed to the next step.
- Interact with the Chatbot:
Run the user interaction script:
python furhat_agent.py
To see a list of dependencies used in this project, refer to the requirements.txt file:
cat requirements.txt