I am a PhD student in the Dialogue Systems and Machine Learning group at Chair for Dialog System and Machine Learning, Heinrich Heine University in Düsseldorf, supervised Prof. Dr. Milica Gasic.

My research focuses on developing emotionally intelligent task-oriented dialogue systems that integrate affective understanding, management, and expression into human-computer interaction. Creating EmoWOZ, the first large-scale emotion-annotated corpus for task-oriented dialogues, I have advanced methods for emotion recognition, natural language generaiton, and multi-turn reinforcement learning using large language models for dialogue policy optimisation. My work bridges affective computing and dialogue system research, aiming to make task-oriented AI communication more human-like, robust, and emotionally attuned.

In addition, I am also interested in other topics such as the modelling of task-oriented modules as well as other aspects of large language models such as prompt optimisation and safety.

This website provides an overview of my ongoing research, publications, and selected projects. Feel free to get in touch via e-mail or LinkedIn in English or Chinese.