Tóth, ErzsébetKhan, Dawood Ahmad2024-06-232024-06-232024-04-13https://hdl.handle.net/2437/374611This thesis delves into the forefront of Natural Language Processing (NLP) by leveraging cutting-edge Transformer models. Through the utilization of technologies like PyPDF2, python-dotenv, faiss-cpu, Streamlit, and Langchain, a robust NLP framework is constructed. Langchain serves as the cornerstone, enhancing both language generation and comprehension. PyPDF2 facilitates seamless integration with PDF documents, while python-dotenv ensures secure management of environment variables. By combining state-of-the-art libraries with transformative AI models, this study showcases the profound impact on document processing, virtual assistants, and beyond, pushing the boundaries of NLP towards more natural and dynamic interactions.47enNatural Language ProcessingPDF ChatBotNatural Language GenerationNatural Language UnderstandingPythonGenerative AIInformaticsHozzáférhető a 2022 decemberi felsőoktatási törvénymódosítás értelmében.