NaCTeM

Professor John Keane

Position: Deputy Director

Telephone: +44 161 306 3334
Email: John.Keane@manchester.ac.uk

Profile

John Keane holds the MG Singh Chair of Computing Science and leads the internationally renowned Data and Decision Engineering group. His primary research activity is at the confluence of high performance computing and data mining. His interests relate to the design, development, verification, and engineering of algorithms for data-intensive systems. He has edited one collection of papers, and is co-editor of 3 volumes in the Springer-Verlag Lecture Notes series.

Keane’s work (as Principal or Co-Investigator) has received funding of over £1M from UK and European funding bodies over the last 8 years. He is currently PI on the ESPRC Distributed Information Management Information Broker for Heterogeneous Information Sources (IBHIS) project, part of the EPSRC DIMNet project, CI on the BBSRC Protein Functional Classification using Text Data Mining project, and PI and CI on two DTI-funded TCS projects on text mining and data fusion respectively. He was previously PI on the EPSRIT High Performance Banking (HYPERBANK) project and CI on the EPSRC Parallel Temporal Theorem Proving (PTTP) project. He was Organisation Co-Chair for COOPIS99, Workshop/Tutorial Chair for EUROPAR01, and was on the Programme Committee of IDEAL02, EUROSIM2001, and BNCOD2001.

Keane was an invited speaker at the EU Workshop on Data Mining and High Performance Computing in 1999, and has been an invited visitor at Hitachi Research Laboratories in Japan. He has been an Executive Director of the internationally renowned Centre for Novel Computing, at the University of Manchester, since 1994, and was a member of the UK CPHC Research Strategy Group. Keane has worked in industry with TSB, Philips, and ICL, and has led technical consultancies for a number of organisations, such as Fujitsu, NATWEST, MTI, DATEL and Minerva Softcare, Belgium.

Research

Professor John Keane investigates correctness and performance issues of concurrent/parallel processing for applications involving large complex data sets. Application areas include commercial domains including banking and retail, and symbolic domains such as theorem proving. Technical issues of concern are distributed querying, language and system models, data mining, high performance parallel systems, and reasoning and design for distributed algorithms. His research projects have been funded since 1994 by the EPSRC and the ESPRIT programme. Current research students are investigating association rules (interesting-ness and performance): parallel attribute-oriented induction; parallel graph processing for link analysis; and co-ordination frameworks for distributed financial processing.

School home page