July 2, 2015 - 10:00am PDT
New Challenges in Big Data: Technical Perspectives - Hwanjo Yu
Abstract
Big Data has been a buzz word as an emerging technology. Yet there seems no killing application and it is unclear what new challenges are. One main reason is that Big Data is a methodological discipline applied to a wide range of domains, and its individual application may not look technologically substantial, though it improves the quality of services. However, Big Data technology has its own technical depth and nontrivial challenges. In this talk, we first clarify new challenges in Big Data processing that are distinct from conventional parallel processing. After that, we introduce several research projects in the data mining lab at POSTECH, including PubMed relevance feedback search, blackbox video search, novel recommendation, and timing when to recommend.
Bio
Hwanjo Yu received his PhD in Computer Science at the University of Illinois at Urbana-Champaign in June 2004 under the supervision of Prof. Jiawei Han. From July 2004 to January 2008, he was an assistant professor at the University of Iowa. He is now an associate professor at POSTECH (Pohang University of Science and Technology). He developed influential algorithms and systems in the areas of data mining, database, and machine learning, including (1) algorithms for classifying without negative data (PEBL, SVMC), (2) privacy-preserving SVM algorithms (PP-SVM), (3) SVM-JAVA: an educational java open source for SVM, (4) RefMed: the relevance feedback search engine for PubMed, and (5) TurboGraph: a fast parallel graph engine handling billion-scale graphs in a single PC. His methods and algorithms were published in prestigious journals and conferences including ACM SIGMOD, ACM SIGKDD, IEEE ICDE, IEEE ICDM, ACM CIKM, etc.