Question-Answering-System 📖

About

The Question Answering System is an AI-powered application designed to provide accurate and timely responses to user queries. Built on advanced natural language processing and machine learning techniques, it enables users to obtain relevant information from a large corpus of documents or knowledge bases. Whether you are a researcher, developer, or knowledge seeker, this system offers a powerful tool for efficient information retrieval and decision-making.

Before getting started:

🍴 To incorporate this repository into your own project, you can fork it.
⭐ In order to receive notifications about future improvements, please consider starring this repository.

For a short write up check out this medium post.

Features

Upload documents (PDF, DOCX, TXT) and answer questions about them.

Live Demo

Available on Streamlit

How It Works

1. A preprocessing step is performed to clean and prepare the text data.
2. The documents are split into smaller sections or splits of a specified size.
3. The splits obtained from the previous step are stored, often using a vector store or a similar storage mechanism.
4. The question and answering application retrieves splits from the storage based on certain criteria.
5. An LLM, which refers to a language model, is utilized to generate an answer using a prompt that includes both the question and the retrieved splits.

Running the Program

1. Clone the Repository: Start by cloning the GitHub repository containing the question and answering system.
2. Navigate to the cloned repository directory and install any necessary dependencies or packages.
3. Check if the question and answering system requires any specific configuration. Look for configuration files, such as config.yaml or settings.py, and modify them if necessary.
4. Run the command to start the question and answering system.
5. Once the system is up and running, you can interact with it by sending questions and receiving answers.

Related Material

For more information and resources on building and improving question answering systems, refer to the following:
Research Papers on Question Answering Systems
Natural Language Processing (NLP) Tutorials and Courses
Coursera Courses
How I Built A Question-Answering System for Architects

License

See the LICENSE file for license rights and limitations (MIT).