Goal-setting dialogue for physical activity with a virtual coach

This github repository contains the code for the chatbot Jody that is created for the thesis project: Goal-setting dialogue for physical activity with a virtual coach. Please refer to our OSF pre-registration for more details on our experiment.

Dialog flow

The figure below visualizes the structure of the dialogue with Jody.

System architecture

Frontend

The frontend is a html-page that makes use of Rasa Webchat 1.0.1.

It is expected that the user ID is provided as a URL-parameter, e.g. http://<IP_address>:5005/?userid=beyza if the frontend is running on port 5005. This user ID is extracted and sent to the backend to link previous collected data to that user.

Files:

• index.html: html-page if the conversational agent runs locally.
• socketChannel.py: This file is needed to connect the frontend to the backend.

Backend

The main component is a conversational agent trained in Rasa 2.8.0.

Files:

• actions: custom actions, e.g. to read from files.
• models: contains trained models.
• config.yml: configuration for the training of the agent.
• data: contains files to specify e.g. the stories on which the agent is trained.
• domain.yml: utterances, slots, forms etc.
• endpoints.yml: defines the endpoints of the conversational agent.

Experiment data

The experiment_data folder contains the following two files:

• goals.csv: contains the examples of people that achieved a running or walking goal.
• user_examples: this file contains example usernames that you can use to chat with Jody.

Running Jody

To run the conversational agent locally:

1. Install the python package Rasa 2.8.0.
2. In a command window, navigate to the root folder and type rasa run -m models --enable-api --cors "*".
3. Open a separate command window and type rasa run actions to start the custom action server.
4. Open the frontend ("index.html") and specify a userid in the URL. Choose one of the usernames that can be found in /experiment_data/user_examples. For example: index.html?userid=beyza
5. Chat with the conversational agent.

License

Copyright (C) 2022 Delft University of Technology.

Licensed under the Apache License, version 2.0. See LICENSE for details.