Lizi Liao     

Assistant Professor
School of Computing and Information Systems
Singapore Management University
  lzliao at smu dot edu dot sg
Office: SCIS2-4056
[Google Scholar]

I am a faculty member of the School of Computing and Information Systems at SMU. My research explores two questions: What are the underlying principles of humans understanding conversation context as well as making proper responses, and how we can implement them on machine learning models? Research on this topic has to necessarily be at the intersection of Machine Learning, Natural Language Processing and Multimedia. In my lab, we are specifically interested in task-oriented dialogues, proactive conversational agents, and multimodal conversational search and recomendation as the application target. I received my Ph.D. from National University of Singapore, advised by Professor Tat-Seng Chua.

  I'm recruiting 0-2 new PhD students every year (apply to PhD program and list me as a potential advisor). Our group also has multiple positions for summer interns and visiting research students. Please feel free to email me with your CV if you are interested.
What's New
Research Highlights

Proactive Conversational AI

We recently published one of the earliest works on developing proactive dialogue systems in the era of LLMs [EMNLP'23a]. To improve the proactiveness of conversational agents, we research on automatic ontology expansion [EMNLP'23b, EMNLP'22a], target-driven conversational recommendation [EMNLP'23c] and building unified user simulators for better support [EMNLP'22b]. We also actively organize tutorials about proactive conversational agents [WSDM'23, SIGIR'23] to discuss important issues in conversational responses’ quality control, including safety, appropriateness, language detoxication, hallucination, and alignment.

Multimodal Conversational Search and Recommendation

Search and recommendation systems prevail and have profound impact. We aim to bridge the information asymmetry problem between the user and system via multimodal conversation [MM'18]. It centers on broader types of ‘understand’ the user and ‘respond’ to the user under certain context. Specifically, we look into multimodal dialogue understanding [MM'23a], state tracking [TMM'22], knowledge-aware response generation [SIGIR'22a] and response strategy modeling [MM'22].

Conversation AI + X (ChatPal, Learning Companion) Interdisciplinary Research

We work on a range of interesting and useful applications that aims to improve human life and society with conversational AI. A line of our research has focused on utilizing LLMs as "teachers" to enhance smaller models in various tasks such as emotional support [arXiv'23, MM'23a]. We also recently started to develop question generation models for online learning companion. We also take a great interest in multimodal data, including work on human-in-the-loop video monent retrieval[MM’23b] and e-commerce data towards intelligent shopping assistant [SIGIR’22b].

Current Offering:
Previous Offerings:

Organizing Committee:
Senior Programe Committee Member:

When I have spare time, I enjoy reading books, hiking, and dancing.

Webpage template borrowed from Prof. Wei Xu.