Research

Our group develops scalable systems for querying and mining large datasets, explores ways to compress and use AI models efficiently, and applies data-driven methods to problems across different domains.

Scalable Data Mining Systems

Develop data structures and algorithms for scalable querying, retrieval, and pattern discovery over large and complex datasets
Design streaming and adaptive computation for continuously evolving data and changing workloads

Selected related publications

Model Efficiency & Compression

Compress models through pruning, quantization, distillation, and low-rank or tensor-based methods
Enable efficient adaptation and inference for modern AI models, including foundation models

Selected related publications

AI + X: Data-Driven Problem Solving

Apply AI and data-mining methods to address problems in a field of interest
Combine domain knowledge with computational methods to analyze data, discover patterns, or make predictions

Selected related publications

These directions are connected by a shared interest in efficient, scalable data analysis. They are examples rather than fixed project boundaries, and a project may span more than one direction. See all publications.