Thursday, March 16, 2006

Event: What is e-Social Science?

Another Access grid seminar today, this one very much an advert for the National Centre for e-Social Science's products, which are strangely parallel to Tim Berners-Lee's. in fact, I wonder if they even know about one another. The main difference between the Access Grid/NCeSS and the Semantic Web is that the latter is free to access whereas the former is a gated proposition.

Anyway, Rob Proctor, Research Director of the NCeSS discussed e-Science, a globally connected, scholarly communit promoting the highest quality scientific research. The aim is to create a global infrastructure for global collaboration (thus my mental vision of its proximity with the SW). Within e-Science are all of the technologies which allow all scientists to collaborate in new ways, necessary because these tools help to make progress in new scientific directions.

The drivers of this programme are the movement towards multidisciplinarity and the data deluge (volumes of data which are unprecedented in research, which we must figure out in order to gain knowledge and understanding of social phenomena). Particularly in reference to the latter, the example he used was that The Bible is around 5MB of data, whereas an Internet Archive from 1996-2002 has 100 TBytes (a thousand million more times than 1 MB - wow, I'll have to start using another hyperbole from now on). That's a lot of data, and we need tools to help us with this deluge.

One example he gave was Discovery Net, which is a data mining and text mining tool. it gets through large amounts of data by looking for canditate ideas which can be modified and tested in new areas. It looks for connections no one's noted because there's too much data to get through. It represents a shift in research approaches, though, as it's data-driven rather than hypothesis-driven.

e-Science is multi-disciplinary, inter-disciplinary, large-scale, data intensive (rather than hypothesis intensive - thus starting a "a different stage in research"), an computationally intensive.

It's all connected through "The Grid", an infrastructure which includes "not only [the use of others'] computers but also data storage resources and spcialised facilities" which may be proximally distant. It's software which is transparent and scaleable.

The Access Grid is an example of one of the grids used for e-Science. It allows distributed researchers to collaborate in virtual research environments in order to overcome the limitations of distance. There are other grids, including the computational grid (e.g., the National Grid Service for crunching and processing, like a distributed power source), a data grid (to generate and access heterogenous data) and a sensor grid (to aid in collaborative collection).

e-Social Science drivers include the movement towareds evidence-based policy making (to asnwer more challenging research questions), interdisciplinary collaboration, and the desire to make better use of existing and new data.

Essentially, we need to be able to exploit text and data mining techniques iin order to find new questions and integrations.

Key to future research through the NCeSS are applications of e-SS (i.e., investigating the use of the grid technologies for tackling substantive research problems, like multiple datasets which are complementary taken in collaboration, but alone have no real depth or bearing) and promoting interdisciplinary research.

So, like I said, an ad.

0 Comments:

Post a Comment

<< Home