Supplied by high-throughput tools and technologies, massive amount of data is being collected in many sectors of our society. Such data holds the promise of supporting ground-breaking discoveries and critical decision-making. Our research is to use and develop powerful analytics tools to unleash the potential of this wealth of data to improve the way in which we understand and interact with the social, technological and business world around us.
Our research follows the philosophy of “from practice, to theory and back to practice”. We look to the real world for practical and important problems; then, we design solutions with provable properties and solid theoretical backing; finally, we build and deploy real systems based on these solutions. Our approach is multidisciplinary by nature, using “tools” from algorithms, machine learning, computional sociology and business analytics.
Our current research focuses on the following three thrusts.
Thrust 1: Information Privacy and Ownership
The concerns about information privacy and ownership are unprecedented in this era. Through linking data from different sources, even disclosing seemingly harmless information may incur significant privacy risks. Our first research thrust focuses on countering privacy threats in varied data-intensive systems, attempting to answer the critical question of “how to achieve sufficient privacy protection, yet at minimum cost?” Towards this, our work focuses on analytically modeling the intricate interplay between privacy protection, information utility and system overhead and optimizes these metrics under user-specified requirements. We have designed and implemented a suite of tools to address practical privacy concerns in a range of domains, from data mining systems to eHealth applications to mobile location-based services.
Thrust 2: Security Threat and Risk Assessment
As the transformative technology for security analyses, data-driven security enables analysts to fuse heterogeneous information from a multitude of sources and vet security incidents from multifaceted perspectives. Our second research thrust focuses on building analytical tools that leverage this capability to support accurate and timely assessment of cyber-security threats and risks, with applications including: more comprehensive profiling of device, user and application behaviors, more accurate risk estimation for access control, better use of multiple security defense systems, and design of tools helping people better manage their data.
Thrust 3: Network Science
Besides tackling concrete privacy and security challenges, we are also interested in developing intelligent solutions with potential to address a range of cross-domain problems. Towards this end, our third research thrust is concentrated on addressing questions repeatedly encountered in the attempt to understand complicated social and technological systems, such as finding micro-level root cause of macro-level phenomena, understanding and modeling network influence at microscopic scales, and aggregating information from interdependent parts of complex systems.