Currently I'm a staff engineer in SAP, designing query processing modules for Sybase ASE, Sybase's flagship product.
Priviously I was associated with the DBS Kernel R&D department of Teradata, and designing the database kernel for Teradata DBMS, the world's leading data warehouse system.
Before that I obtained my Ph.D from North Carolina State University, with a co-major in Computer Science and Operations Research. During my PhD research I did extensive joint analysis of the security, privacy and utility for large dataset in various contexts including relational microdata, social network and sensor network intrusion detection data. I've also interned at AT&T Shannon Laboratory doing research on graph database privacy.
Query Processing, Sybase ASE
Working on query optimizer design and query execution for Sybase ASE, Sybase's flagship OLTP RDBMS that powers Wall Street.
Feature branch manager and feature developer for the Row Level Security (Mandatory Access Control) feature team.
I work in the details of all database internal components in the Teradata Parser and AMP modules. Specifically I¡¯m responsible for
(1) Design, coding, and testing feature in database internals, including query parsing and resolving, query optimization, query plan generation and task dispatching, logon control and session control, trusted session setup, local/global cache system design, data backup and restore, transaction processing, etc. Extending feature for SQL modules as DDL/DML/DCL, stored procedures, access logging etc.
(2) Managing the feature branch repository and version control, coordinate D2D development/testing with other Teradata database/client utilities departments, and maintaining product builds on multiple Linux and windows platforms.
(3) Providing code review and design/testing document review.
Provide maintenance fixes, and internal/customer support in database kernel for the existing supported releases.
Microdata protection
(1) Design novel schemes for the anonymization of relational microdata, which greatly enhance the usability of anonymized data while satisfying the standard privacy protection goals. Propose advanced privacy goals for numeric microdata, and optimize the anonymization algorithms to provide accurate query answering.
(2) Design anonymization schemes that disguise the distribution of sensitive attributes of microdata. Design optimization algorithms to accurately answer aggregate queries. The designed schemes also support flexible, user-defined tradeoff between privacy and data utility.
Wireless Sensor Network Security
Develop an application-independent framework to identify compromised nodes in sensor networks. Design a graph-based reasoning algorithm that can accurately identify compromised nodes even when they collude or form a local majority.
Reputation-based trust management.
Develop a general framework that classifies existing trust functions for reputation-based trust management systems, and propose a statistical method to do behavior profiling and filter out dishonest players in reputation systems.
Analyze the privacy implications of bipartite graph data (e.g. routing table, customer shopping record, etc.). Design anonymization schemes, and evaluate their privacy as well as usability through theoretical analysis and experimental validation on large-scale (Gigabyte) real datasets.
Distribution-based Microdata Anonymization (with Nick Koudas, Divesh Srivastava, Ting Yu)
in the 35th International Conference on Very Large Data Bases (VLDB), 2009. (acceptance rate: 17.9%)
Anonymizing Bipartite Graph Data using Safe Groupings (with Graham Cormode, Divesh Srivastava, Ting Yu), in the 34th International Conference on Very Large Data Bases (VLDB), 2008. (acceptance rate: 16.5%)
A Framework for Identifying Compromised Nodes in Sensor Networks (with Ting Yu, Peng Ning) in ACM Transactions on Information and System Security (TISSEC), Vol. 11, No. 3, pages 12:1--12:37, March 2008. (Invited submission)
Aggregate Query Answering on Anonymized Tables (with Nick Koudas, Divesh Srivastava, Ting Yu) 23rd IEEE International Conference on Data Engineering (ICDE), April 2007 (acceptance rate: 18.5%)
Patents
"Method and Apparatus for Providing Anonymization of Data", with G. Cormode, D. Srivastava, T. Yu, US Patent 20110041184
"Systems and Associated Computer Program Products That Disguise Partitioned Data Structures Using Transformations Having Targeted Distributions", with N. Koudas, D. Srivastava, T. Yu, US Patent 20100114840
