Chatbot

Our AI team focuses on cutting-edge artificial intelligence research and applications.

Introduction​

Chatbot systems based on Retrieval-Augmented Generation (RAG) are an emerging research direction that aims to build reliable, knowledge-grounded conversational agents by combining information retrieval with large language models. With recent advances in LLMs and vector search, RAG-based chatbots have demonstrated strong potential in real-world applications that require factual accuracy, domain knowledge, and controllable behavior. Typical application domains include education support, public administration services, enterprise knowledge assistants, and human resource management systems.

Our research group focuses on developing end-to-end RAG-based chatbot systems, covering system design, optimization, and comprehensive evaluation. We study techniques for document processing, search and retrieval optimization, and retriever–generator integration to improve response quality and faithfulness. In addition, we build and deploy domain-specific chatbot systems for education, administrative services, and human resources, while establishing practical evaluation frameworks to assess retrieval effectiveness, grounding accuracy, and overall system performance.

Contact: Dr. Nguyen Thi Thu Trang | ✉️ [trangntt@soict.hust.edu.vn]

Research Directions​

• RAG System Design and Optimization: Optimizing end-to-end RAG pipelines, including document processing, chunking, embeddings, and retriever–generator integration.

• Search and Retrieval Optimization for RAG: Improving dense, sparse, and hybrid search, along with reranking and query reformulation for higher retrieval quality.

• Evaluation of RAG-based Chatbots: Developing practical evaluation methods for retrieval accuracy, grounding faithfulness, and end-to-end chatbot performance.

Team Members​

Le Dai Viet

Team Leader

Ho Duc Tu

Researcher

Dang Kim Ngan

Researcher

Le Hai Anh

Researcher