Dates:

• 14.04.2026 - 21.07.2026

Place and Time:

• Lecture: Tuesdays 14:30-16:00; Room: 25.22.U1.34 (start of first lecture: 14.04.26)
• Exercise: Mondays 10:30-12:00; Room: 25.22.U1.34 (start of first exercise: 20.04.26)

Lecturer:

• Dr. Nurul Lubis (lubis(at)hhu.de)

Study programme:

• Master's programme in Computer Science (PO 2015, 2005)
• Master's programme in Artificial Intelligence and Data Science (PO 2019, PO 2021)

Credit Points:

• 5 CP (2+2 SWS)

Language:

• English

Content:

• Introduction: Architecture of a spoken dialogue system, dialogue acts, turn management issues.
• Semantic decoding: representing and decoding meaning from user inputs, semantic decoding as a classification task, semantic decoding as a sequence-to-sequence learning task.
• Dialogue state tracking: tracking beliefs over multiple turns, classical generative and discriminative approaches, recent deep learning approaches, integration of decoding and tracking.
• Dialogue Management: modelling via Markov Decision Processes, reinforcement learning, gradient methods, lifelong and meta learning.
• Response Generation: template methods, generative models, recent neural network approaches.
• Speech interface: automatic speech recognition and speech synthesis.
• Advance topics: End-to-end dialogue systems, large language models (LLMs) and reinforcement learning with human feedback (RLHF), autonomous agents, and affective dialogue systems.
• Students will be provided with a set of Python tools which will enable them to configure and test a simple spoken dialogue system. They will be asked to implement parts of dialogue system modules related to training, design, and evaluation. Students will also implement a reinforcement learning algorithm and optimize the dialogue manager in interaction with a simulated user. This will give them an opportunity to explore a practical example of reinforcement learning.