The importance of data and knowledge extraction in science is growing. Fields as diverse as bioinformatics, geophysics, astronomy, medicine, meteorology and particle physics are among many facing the situation where data are becoming larger and more intricate. Furthermore the requirements scientists have of the data are becoming increasingly complicated, utilizing state of the art analysis techniques such as those made available by the machine learning/data mining communities and increasingly more powerful visualization techniques. Improvements in computational infrastructure make the results from an ever larger number of data and computing resources accessible from the scientist's desktop, even when those resources are distributed worldwide. A prime challenge for e-science is to enable the effective extraction, integration, analysis and presentation of knowledge from the data avalanche, so that the scientist can exploit its potential.
In this working meeting, we identified research and infrastructure directions that the participants considered as important for future funding cycles in e-science. This included work which:
Meeting this challenge effectively will involve the coordinated efforts of specialists in a number of areas, and this workshop brought together people from a variety of backgrounds, to define the problems to be solved and to identify the first steps towards solving them. The programme had the following structure:
10.00 Welcome & intro to NeSC/eSI (Malcolm Atkinson)
10.15 Intro to workshop and its aims (Bob Mann)
10.30 Scientific Motivation overview (Roy Williams)
11.00 Coffee
11.30 Data Mining overview (Chris Williams)
12.00 Data Visualization overview (Ken Brodlie)
12.30 Charge to the Breakout Groups
13.00 Lunch and Breakout groups
15.00 Tea
15.30 SDMIV Application Talk 1 (Bob Nichol)
16.15 SDMIV Application Talk 2 (Jim Austin)
17.00 SDMIV Application Talk 3 (David Gilbert)
17.45 Formal close of Day One
19.30 Dinner
09.00 SDMIV Application Talk 4 (Julian Gallop)
09.20 SDMIV Application Talk 5 (Amos Storkey)
09.45 Current computational infrastructure
10.30 Coffee
11.00 Future computational infrastructure
11.45 Report back from breakout groups and discussion of how current
and future computational infrastructure meets requirements they
have identified
13.00 Lunch
13.45 Choice of topics for discussion in break-out groups, to identify in
up to five areas concrete suggestions for things to do in the
next twelve months.
14.00 Start of breakout sessions
15.00 Tea
15.30 Report back from breakout sessions and panel discussion of group
leaders to produce prioritized programme for work in next twelve
months.
17.00 Close of workshop
The workshop was for those who wished to influence the UK e-science agenda of the future: members of the data integration, machine learning/data mining and visualization communities, computer scientists designing and building the infrastructure these disciplines use and researchers from any data-intensive field who realize that they will need the expertise from all these other areas to do their science effectively.
The workshop was be hosted by the e-Science Institute in Edinburgh which is a centre for education and research for e-science, and provides new state-of-the art facilities including an Access Grid system. All the costs of the workshop, lunches, refreshments and the workshop dinner were covered by the National e-Science Centre.
The e-Science Institute is at
15 South College Street,
Edinburgh,
Scotland.
It is less than 15 minutes walk from Waverley rail station, and from St Andrews square bus stations. It is approximately 20 minutes by taxi from Edinburgh airport (40 minutes by bus). A map and contact information is available here.
Many of the talks from the workshop, and other reports can be found at the NeSC event materials site.