Question-Answer Website

This project is a fully functional question-answer web application that allows users to upload a file and ask questions about its contents. The system processes the file, extracts the text, and enables users to query the text for specific information.

How to run sever

git clone https://github.com/your-username/question-answer-website.git
cd question-answer-website
poetry install
poetry run python main.py

Features

• File Upload: Users can upload various document formats.
• Natural Language Processing: The system processes uploaded documents and answers user questions based on the content of the file.
• Asynchronous Web Server: FastAPI provides a fast, asynchronous backend to handle file uploads and question-answer logic.
• ChromaDB: Used to store and index text for efficient querying.
• Document Parsing: Utilizes PyPDF to handle PDF files and Python-magic to identify file types.
• Language Processing: Leverages LangChain and NLTK for NLP capabilities.

Technology Stack

• FastAPI: Provides the web framework for handling requests and delivering responses asynchronously.
• LangChain: Chains together different language models to help answer complex queries.
• ChromaDB: Efficient vector database for text indexing and retrieval.
• PyPDF: Handles the parsing and extraction of text from PDF files.
• Python-magic: Identifies file types for proper handling.
• NLTK: Aids in text tokenization and processing for natural language queries.

How It Works

• Upload a File: Users upload a file through the frontend interface.
• File Parsing: The backend parses the file to extract the text using PyPDF or another parser based on the file type identified by Python-magic.
• Text Storage: The extracted text is stored in ChromaDB, allowing for fast, efficient searches.
• Ask a Question: Users can ask questions related to the content of the uploaded file.
• Answer Generation: LangChain processes the user's query using NLP techniques, returning the most relevant answer from the document.