Dr. Shuai Wang (Dylan)

Dr. Shuai Wang (Dylan)

1113 citations · h-index 15 · 01/05/26

Postdoc, University of Queensland; PhD (2025); Previously: Naver Labs Europe, Leipzig University

Dr. Shuai Wang (Dylan Wang) is a Postdoctoral Research Fellow at the IELab, University of Queensland, Australia. His research lies in Information Retrieval (IR), Natural Language Processing (NLP), and Retrieval-Augmented Generation (RAG), with a focus on building efficient, reliable, and reproducible AI search systems. Dr. Wang has authored 20+ peer-reviewed papers in top-tier venues including SIGIR, WSDM, WWW, ECIR, EMNLP, and EACL, accumulating 1113+ citations with an h-index of 15. He completed his PhD at the University of Queensland in 2025 and currently serves as Course Coordinator and Lecturer for INFS7410 (Information Retrieval and Web Search) at UQ, and as the Communications Chair for SIGIR 2026. Representative work includes COCOM (a 5.69× RAG inference speedup), AutoBool (reinforcement-learned Boolean query generation for medical systematic reviews), and BERGEN/FeB4RAG (reproducible RAG evaluation tooling).

👋 Welcome!

I’m Shuai Wang (Dylan), a Postdoctoral Researcher at IeLab, UQ working on efficient LLM and retrieval systems. I build search and RAG methods that make AI systems faster, cheaper, and easier to evaluate, especially for evidence-based medicine and other high-stakes information work.

I completed my PhD under the guidance of Professor Guido Zuccon, Associate Professor Bevan Koopman, and Dr. Harrisen Scells.

🎓 Academic Background

  • Doctor of Philosophy - University of Queensland (2025) Read My Thesis
  • Master of Engineering Science - University of Queensland (2021)
  • Bachelor of Science - University of Western Australia (2019)

🔬 Research Focus

My work connects information retrieval, NLP, and AI systems:

  • Efficient RAG and LLM inference: context embeddings, KV-cache/memory optimisation, and unified retrieval-generation representations
  • Evidence-based medicine search: MeSH suggestion, screening prioritisation, seed-driven methods, Boolean query generation, and clinical question answering
  • Evaluation infrastructure: benchmarks and open-source tools for RAG and federated search, including BERGEN and FeB4RAG
  • Adaptive retrieval and ranking: 2D Matryoshka retrieval, prompt variation studies, and robust LLM-based rankers

👨‍🏫 Teaching & Mentoring

I serve as Course Coordinator and Lecturer for INFS7410 (Information Retrieval and Web Search) at UQ, teaching 120+ Master’s students across classical IR, dense retrieval, LLMs for search, RAG, and evaluation. Previously, I tutored INFS7410, INFS7205, and DATA7901/7902/7903.

I mentor students on retrieval, RAG, and biomedical NLP projects; several have progressed to PhD scholarships and SIGIR/ECIR papers. If you are interested in research collaboration, please reach out.

🌍 Industry Experience

Research Intern at Naver Labs Europe (Feb-July 2024). I co-led COCOM from method design and implementation to experiments and WSDM 2025 publication, reducing RAG inference cost by 5.69x at near-baseline answer quality.

💼 Job Opportunities

Since February 2025, I have been working as a Postdoctoral Researcher at UQ. I am open to academic and industry roles where efficient LLM systems, search/RAG, trustworthy evaluation, or biomedical evidence technologies are central. If this matches your team, please feel free to reach out.

📰 Latest News

Celebrate

Three Papers Accepted in SIGIR 2026

[Reproducibility Paper]: Beyond Chunk-Then-Embed: A Comprehensive Taxonomy and Evaluation of Document Segmentation Strategies for Information Retrieval; [Reproducibility Paper]: The Vulnerability of LLM Rankers to Prompt Injection Attacks; [Short Paper]: Rank-R1: Enhancing Reasoning in LLM-based Document Rerankers via Reinforcement Learning.

Travel

Serving as Communication Chair for SIGIR 2026

I will be serving as the Communication (Social Media) Chair for SIGIR 2026 conference to be held in Melbourne | Naarm, Australia.

Celebrate

Paper accepted in EACL2026 Main Conference

[Full Paper]: AutoBool: Reinforcement-Learned LLM for Effective Automatic Systematic Reviews Boolean Query Generation.

Read All News

🤝 Professional Services

I contribute to the academic community by serving as a PC/SPC (reviewer) member for:

📚 Journals

  • TOIS: ACM Transactions on Information Systems
  • Journal of Data and Information Quality

🏛️ Conferences

  • ACM ICTIR 2023, SIGIR 2024, SIGIR 2025, SIGIR 2026
  • ECIR 2024, 2025
  • WSDM 2026

📝 Publications

My publications are authored under the name Shuai Wang. As this is a common name in academia, please refer to my ORCID (0000-0002-0726-5250), Google Scholar, or Semantic Scholar profile to disambiguate my work from other researchers with the same name.