Yun Chi Research Staff Member NEC Laboratories America, Inc. 10080 North Wolfe Road, Suite SW3-350 Cupertino, CA 95014-8505 Emails: Phone: (408) 863-6044 Fax: (408) 863-6099

Quick Links:    Research  |   Publication  |   Software  |  

Biographical Sketch

Research Interests

• Cloud Computing

  Apply machine learning, optimization, control theory to problems in cloud computing. Answer questions such as: How to efficiently schedule and dispatch jobs given historic query loads? How to accurately estimate the execution time for a database query? How to design effective simulations for capacity planning? How to conduct cost-sensitive learning in cloud computing?

• Social Networks

  Analyze social networks as dynamic graphs. Study communities, their main topics, and their evolutions. Find special communities such as spam farms. Rank community members based on their influence. Make recommendation based on historic behavior. Answer questions such as: How to extract meaningful communities based on links and contents? How to detect main topics and their trends in the communities? How to quantify, detect, and visualize the evolution of communities? How to tell if a community member has large amount of influence? How to make profits from community members and their influences?

• Machine Learning Applied in Databases

  Identify main data dimensions that are common in most of networked data. Introduce structures into unstructured data, using machine learning techniques. Fuse multiple features for a learning task such as classification. Update learning results in an incremental fashion as new data arriving constantly. Answer questions such as: How to introduce structures into networked, time-varying, multimedia data, such as blogs? How to meaningfully combine heterogeneous information such as links, contents, and demographic data in the learning process? How to conduct OLAP-type operations on the introduced structure? How to efficiently handle information retrieval on the above data and structure? How to incrementally incorporate new data by updating previously learned results?

• Mining Patterns in Structured Data

  Mine structured data such as trees and graphs. Apply the mining algorithms to discover useful patterns in social networks. Answer questions such as: What kind of tree and graph patterns are interesting? How to quickly find patterns from huge social networks? How to obtain useful insights from the mined patterns? How does information propagate in a social network? How to efficiently unify global features with local (personal) features? How to extract meaningful communities based on links and contents?

Recent Professional Activities

• Program Committee, ACM International Conference on Information and Knowledge Management (CIKM 2008, 2009)
• Program Committee, ACM International Conference on Research and Development in Information Retrieval (SIGIR 2010)
• Program Committee, IEEE International Conference on Data Mining (ICDM 2006, 2009)
• Program Committee, IEEE International Conference on Data Engineering (ICDE 2009)
• Program Committee, SIAM International Conference on Data Mining (SDM 2009, 2010)
• Program Committee, International Conference on Very Large Data Bases (VLDB 2010)
• Program Committee, International World Wide Web Conference (WWW 2009, 2010)
• Program Committee, International Conference on Machine Learning (ICML 2010)
• Program Committee, International Workshop on Mining and Learning with Graphs (MLG 2007, 2008, 2009)

Awards

• Best Research Paper Award Runner-up, KDD 2007
• UCLA Chancellor's Dissertation Year Fellowship, 2005

Last updated: Fri Jan 01 12:38:54 PST 2010
SE: 529_167