Introduction to the Knowledge Discovery Department Institute for Infocomm Research Limsoon Wong Deputy Executive Director (Research) I2R: Imagination to reality I2R • BMRC SERC IMRE IME • DSI IHPC ICES SIMTECH • Advance Infocomm Technology to benefit humanity and create prosperity for S’pore core values: Dedication to excellence, passion for innovation, integrity & respect size: 350 RSEs mission: EDB-iDA ICT Cluster Map Organization I 2R Services & Applications Infocomm Security • Cryptography • Network security • Information security Communications & Devices Knowledge Discovery E-Service Infrastructure • Knowledge Extraction • Decision Systems • Discovery Systems • Service Composition • Service creation • Data-in-Network • Connected Home • Secure Filesystem/ DBMS Embedded Systems Radio Systems • Terminal • Software Defined Radio • Transceiver system • Antenna • Radio Over Fiber • UWB Media Media Processing • Mutimedia Signal Processing • Noise Reduction Networking • Next Gen IP • End-to-end QoS • Mobile Middleware • Mobile Adhoc Networks • Mobile Sensor Networks Media Semantics Human Computer Interaction • Media Adaptation • Mixed Media Mining • Natural Language Synergy • Pervasive Media Digital Wireless • Modem Technology • Multiple Access • Smart Antenna Array Systems • Perceptual Human Media • Speech & Dialogue • Biometrics Lightwave • Network Design • WDM Systems • Fibre Technology Services & Applications Information Security Cryptography Cryptographic techniques for privacy protection, highspeed encryption, group digital signatures and new public key infrastructure Information Security Techniques for privacy protection, peer-to-peer system security and media security Knowledge Discovery Knowledge Extraction Service Creation Data fusion from sensors, knowledge extraction technologies. Applications in functional genomics & gene regulatory networks Policy-based framework for creating/managing application, storage and network services Data-in-Network Decision Systems Data Mining technologies. Applications in clinical data analysis and protein network analysis Discovery Systems Network Security Wireless network security, intrusion detection and technology to trace attackers E-services Infrastructure Data cleansing, modeling & knowledge management technologies. Applications in immunoinformatics and venominformatics High-speed data cache and IPC on optical network Service Composition Tech for consumers to create personalized services from component eservices Connected Home Reference design for an OSGi compliant residential gateway Secure Filesystem Steganographic and reliable file systems/DBMSs on externally managed platforms History 8 years of bioinformatics R&D in Singapore Integration Technology (Kleisli) MHC-Peptide Protein Interactions Extraction (PIES) Binding (PREDICT) Gene Expression Molecular Cleansing & Connections & Medical Record Warehousing Datamining (PCL) (FIMM) Gene Feature Recognition (Dragon) Venom Informatics GeneticXchange 1994 ISS 1996 1998 KRDL 2000 Biobase 2002 LIT/I2R Directions Knowledge Extraction • gene regulatory network • endocrinology info system Translate inspiration from biological systems into advancement in life and computing sciences Discovery Systems • immunoinformatics • venominformatics Advance data mining technologies in decision systems for complex problems Technologies for • data mining • data fusion • data cleansing • data modeling • knowledge extraction • knowledge management Decision Systems • clinical data analysis • protein interaction network Decision Systems Lab Highlight PCL Technology • Machine learning of emerging patterns • Rules learned by computers easily understood by doctors • High accuracy Yeoh et al. Cancer Cell 1, 133-143, 2002 Discovery Systems Lab Highlight FANTOM2 Consortium Nature 420, 563-573, 2002 • Brusic et al. Cytokinerelated genes identified from the RIKEN full-length mouse cDNA dataset. Genome Res. (in press) • Silva et al. Identification of novel human disease-related gene candidates from the RIKEN full-length mouse cDNA dataset. Genome Res. (in press) KE Lab Highlight • Dragon Promoter Finder and Dragon Gene Start Finder • Very high accuracy, order of magnitude improvement • Licensed to Biobase, integrated with TRANSPLORER and TRANSFAC Qualification • 18 full-time staff (9 PhD, 6MSc, and 3 BSc) and 10-20 students at any time • Our faculty holds adjunct professorial positions (SANBI, Univ. of Canberra, NUS, and NTU) • During the last two years 8 MSc students graduated under our supervision Professional Activities • • 3 chief editorial roles, and 9 editorial board memberships International Conferences (last 5 years): 4 conference chairs, and more than 40 program committee roles, 10 keynote lectures, and dozens of plenary talks Journals: Track Record • • • More than 100 in last 5 years, many in top journals (e.g. Nature, PNAS, Cancer Cell, JACM) and top conferences. Patents (last 5 years): 1 granted and 6 pending Commercialization (last 5 years): 2 international start-ups and one software licensed Publications (last 5 years): People Decision Systems Head: Dr. See-Kiong Ng PI: A/Prof. Limsoon Wong, Dr. Jinyan Li Staff: Huiqing Liu, Zhuo Zhang, Soon-Heng Tan Post Doc: Dr. Shaowu Meng Discovery Systems Head: A/Prof. Vladimir Brusic Staff: Judice Koh, Guanglan Zhang, Seng Hong Seah Post Doc: Dr. Rekha Pillai Knowledge Extraction Head: Prof. Vladimir B. Bajic Staff: Hao Han, SPT Krishnan, Allen Chong, Sin Lam Tan Post Doc: Dr. Suisheng Tang Here they are...