Joel Mackenzie

Dr. Joel Mackenzie is a Lecturer at the School of Electrical Engineering and Computer Science at the University of Queensland. His research focuses on efficient and effective representations for large-scale search engines, including indexing, compression, and retrieval. He has published extensively in top-tier information retrieval conferences and journals.

His research interests include:

• Information Retrieval
• Search Engines
• Algorithms and Data Structures
• Indexing and Compression Techniques
• Query Processing and Optimization
• User Behavior Analysis

Mackenzie's research explores both theoretical and practical aspects of search systems, with a particular focus on empirical experimentation and efficiency optimization. His recent work has examined neural information retrieval, dynamic pruning algorithms, and the representation power of modern language models for search applications.

His publications show a strong trend toward optimizing search efficiency while maintaining effectiveness, with recent papers addressing challenges in neural information retrieval, lossy compression for indexes, and measuring search quality through novel metrics. His work spans both traditional information retrieval techniques and cutting-edge approaches leveraging modern language models.

Mackenzie has received research funding from Google Asia Pacific for his project "Compressing Neural Indexes for Short and Long-Term Applications" (2024-2026).

He is actively involved in the information retrieval research community, serving as a program chair for ADCS 2022 and as a short paper co-chair for SIGIR 2025. He has also been a reviewer for numerous top conferences including SIGIR, CIKM, and WSDM.

Dr. Mackenzie supervises several PhD and Masters students working on projects related to next-generation information retrieval systems, online discussion summarization, and human-centered verification of language model outputs.