معرفی
Charles Clarke is a Professor in the Department of Computer Science at the University of Waterloo. His research focuses on information retrieval, search, question answering, data science, and software tools. He holds a PhD in Computer Science from the University of Waterloo (1996), an MMath in Computer Science (1990), and a BSc (Honours) in Mathematics/Computer Science from Memorial University of Newfoundland (1986).
Clarke has organized and contributed to major workshops, including LLM4Eval and Search Futures, which explore the role of large language models (LLMs) in evaluation and future search technologies. His work emphasizes human-AI collaboration, adversarial testing of models, and novel evaluation metrics like Normalized Residual Gain.
- Key contributions include frameworks for benchmarking LLM-based relevance judgments and analyzing prompt sensitivity in AI systems.
- He has published extensively on topics such as adversarial attacks on neural ranking models (EMPRA) and generative approaches to information retrieval evaluation.
- Clarke actively participates in conferences like SIGIR and WSDM, contributing to testbeds and community-building initiatives in IR research.
His research bridges theoretical advancements in AI with practical applications, addressing challenges in search effectiveness, security, and human-AI synergy.
