This project explores using Rasa to build fitness chatbots that provide personalized recommendations based on user data such as height, weight, and fitness goals. By leveraging AI, the chatbot adapts advice to individual users, improving engagement and helping them achieve better fitness outcomes. The research also examines the effectiveness of these personalized interactions and practical challenges in implementing AI-driven customization.
The Bot not only address the user's intent, it also attempt to suggest reasonable suggestion for the user.
Rasa is a framework that makes things eaiser to build custom chatbots. It can give you instights to what people are trying to say and how do the conversations go.
- It is highly customizable and suportive
- It uses NLP and Symbolic rule-based systems
To install Rasa we will definately need to install Python versions between (3.8 to 3.11) For this project, I recommend to install Python version of "3.9.13".
And if you want the rasa project compeletly isolated from the global enviorment, it is a good idea to use virtual enviroment for the system.
It can be easily done using Anaconda.
DIET Classifier
The DIET classifier is Rasa’s integrated machine learning model. Its duty is to preform intent classification and entity extraction. It performs such functionalities by manipulating and utilizing of transformer architecture with the aim of improving accuracy. The enhancement of the accuracy is necessary especially for computationally efficiency. Unlike traditional models, DIET can process both intent and entity recognition simultaneously. As the result, it can reduce dependency on pre-trained word usage. It supports custom datasets and making it adaptable to various chatbot applications.
I recommend using Anaconda for managing Python environments and packages. Anaconda simplifies installation and avoids dependency issues. Steps
- Install Anaconda (Python & R distribution).
- Use Anaconda Prompt (avoid PowerShell).
- Create a new environment (example name: installingrasa)
You can choose a more descriptive name for your environment.
- Activate the environment
- Install required packages
