Shengyao Zhuang

Shengyao Zhuang is a Postdoctoral Research Fellow at CSIRO's Australian e-Health Research Centre in Brisbane and an Adjunct Lecturer at The University of Queensland (UQ) since 2023. His research focuses on large language model-based search systems for medical domains and Natural Language Processing in information retrieval.

• Education: PhD in Computer Science (UQ, 2023), Master of Information Technology (UQ, 2018), Bachelor of Electrical Engineering (Chongqing University of Science and Technology, 2016)

Shengyao's work spans information retrieval, NLP, and zero-shot ranking with large language models. He develops neural rankers, efficient validation toolkits (Asyncval), and hybrid retrieval systems that combine dense and sparse representations. His contributions include contextualized exact term matching (TILDEv2) and counterfactual bias mitigation techniques for implicit feedback.

Recent publications analyze LLM-based stemming, Vec2Text threats to retrieval systems, and green computing impacts of water consumption. His methods improve cross-lingual retrieval and federated learning efficiency while maintaining sub-100ms latency for CPU environments.

• Scientific Awards:
  • SIGIR 2021 Top10 authors (unofficial) Rank #3
  • 2022 HUAWEI DIGIX GLOBAL AI CHALLENGE Champion

As a Tutor at UQ (2019-2022), he taught INFS7410: Information Retrieval and Web Search. He serves on program committees for major conferences including SIGIR, TheWebConf, and ECIR.