FestBot is a web application built on Streamlit that facilitates information extraction and retrieval based on PDF inputs related to festivals or organizations. The application utilizes Retrieval-Augmented Generation techniques and embeddings for efficient query handling.
Features
-
PDF Input Handling: Users can upload PDF files containing information about festivals or organizations.
-
Chatbot Integration: A chatbot component processes user queries and retrieves relevant information from the pdf Input.
-
Delete Chat History: Previous conversations with the ChatBot can be deleted.
Workflow
- Embeddings: Text from uploaded PDFs and user queries is converted into numerical representations called embeddings using natural language processing techniques.
- Vectorstore: These embeddings are stored in a FAISS vectorstore, optimized for fast similarity search.
- Similarity Search: When a user submits a query, FestBot retrieves embeddings of similar content from the vectorstore to generate relevant responses.
- Response Generation: Based on the closest embeddings found, FestBot presents information extracted from the PDFs as answers to the user's queries.
Technology Stack
-
Python: Programming language used for development.
-
Streamlit: Framework used for building interactive web applications.
-
Langchain: Library utilized for natural language processing tasks, including text extraction and embeddings generation.
