Z-ENG: Large Language Models for Intelligent Event Assistance

2026-2027 ősz

Nincs megadva

Téma leírása

Description

This project investigates the application of Large Language Models (LLMs) to support visitors, exhibitors, and event organizers. Students will extend the existing AI assistant by improving document understanding, conversational capabilities, multilingual support, and automated knowledge extraction.

Possible Subtopics

3.1 Advanced RAG-based Event Assistant

Develop an AI assistant capable of answering questions across multiple exhibition stands and information sources.

Tasks

  • Build multi-document retrieval pipelines.
  • Improve document indexing.
  • Evaluate retrieval quality.
  • Reduce hallucinations.
  • Support source citation.

3.2 Automatic Knowledge Base Construction

Develop tools that automatically create stand knowledge bases from documents, websites, brochures, and presentations.

Tasks

  • Implement document parsing.
  • Extract structured information.
  • Generate semantic document chunks.
  • Automate indexing.

3.3 Automatic FAQ Generation

Develop LLM-based methods for generating frequently asked questions from uploaded documents.

Tasks

  • Generate FAQs.
  • Generate suggested chatbot prompts.
  • Evaluate answer quality.
  • Integrate generated FAQs into the platform.

3.4 Multilingual AI Assistant

Extend the assistant to support multiple languages for both visitors and exhibitors.

Tasks

  • Translate documents automatically.
  • Support multilingual conversations.
  • Compare multilingual LLM performance.
  • Evaluate translation quality.

Required Skills

  • Python
  • NLP
  • Large Language Models
  • LangChain/LlamaIndex
  • Vector Databases
  • Prompt Engineering 

Maximális létszám: 4 fő