Breast Cancer Chatbot

This project focuses on developing an AI Chatbot API for patients of Breast Cancer, and aims at answering their questions and concerns regarding the disease.

This project was developed for an internship program at Folio3 inc.

The Questions and Answers data for this purpose was collected from the following website:

https://emedicine.medscape.com/article/1947145-questions-and-answers?reg=1

If you would like to use this project for any other chatbot purpose, you may replace the data with your own .csv file.

Note: Your data must contain the following:

• Questions (Including Different patterns for each question)
• Answers
• Tags (Keyword for the topic)
• (Optional) Link for the Full Source

Features

• Streamlit UI
• FastAPI Deployment

Installation

Install FastAPI

  pip install FastAPI

Install Streamlit

  pip install streamlit

Create a Virtual Environment

pip install virtualenv
python<version> -m venv <virtual-environment-name>
<virtual-environment-name>\Scripts\activate

Install the requirements.txt

pip install -r /requirements.txt

Run the front-end file on cmd

  streamlit run front.py

Run the back-end file on cmd

  uvicorn main:app --reload

API Reference

Train the Model

  GET /api/train

Self Learning (Add Questions)

  POST /api/addQuestion

Request Body	Type	Description
Questions, Index Number	string, int	Required. Update Question and Index

Predict The Answer

  POST /api/predict

Request Body	Type	Description
Question	string	Required. Question to Generate Answer

Authors

• Abdullah Wahid
• Ahmed Ashraf
• Neeraj Prakash
• Hasnain Ali

Under the Supervision of

• Sanif Ali Momin

Used By

This project is used by the following company:

• Folio3

Screenshots