'Me-Science' the New e-Science

"It's not 'e-science,' it's 'me-science.'" This was the conclusion of Carole Goble from the University of Manchester, having given one of the most popular and well-attended keynote speeches of the EGEE'06 conference.

Having worked closely with scientists, specifically in life sciences, Goble said that our understanding of the true motivations of scientists is essential for the development of Grid technologies. "We're not building infrastructures for e- science, we're building them for e-scientists," she explained.

Goble presented the controversial viewpoint that while scientists are motivated by advancing the frontiers of science, this is not necessarily the principal motivation. Instead, she suggested, they are motivated in large part by getting the best results first, in order to gain, among other perks, more funding.

This "selfish scientist" concept, a term coined by Robert Stevens from the University of Manchester, suggests that scientists are not eager to openly distribute their data around Grid communities, which contrasts somewhat to the Grid idea of "sharing." Goble quoted Mike Ashburn from the University of Cambridge when she said, "Scientists would rather share their toothbrush than their data." She views scientists as being altruistic, but protective. "They are happy for other people to have their data and share their workflows, but only after they've had it published."

"We must therefore enable scientists to publish data at the right time in their lifecycle," insisted Goble. To do this, she suggested paying more attention to when people share data and who gets to see it, so that scientists can preserve privacy and control over their work.

Goble is currently developing a project entitled "myExperiment" to attend to this problem. myExperiment is a collaborative research environment that has been influenced by social networking programs such as Wired and Flickr, and is based on the mySpace infrastructure. myExperiment creates an environment for scientists to adopt Grid technologies, where they can define when they share data, with whom they share it and how much of it can be accessed.

Goble also highlighted the potential of tagging the data, similar to the process of social bookmarking currently used to catalog photos on Flickr. This would allow simple cataloging of data so that there is no need to use "command lines," which, she laughed, "can make your eyes glaze over."

The myExperiment interface would be specific to each institution that uses it and may mean that scientists won't notice the difference between the Grid and the Web. "In the same way that you don't see any HTML protocol when searching for the Radio Times, you will not need to see any of the Grid processes working on your data," said Goble.

Goble's speech evoked a good deal of discussion throughout the conference, and the developments of myExperiment will be followed with great interest.

---

Adoption of Grid Technologies by Enterprise

This year's EGEE'06 differed from previous conferences by introducing an innovative business track, providing industrial organizations with a chance to explore how they may take advantage of the grid's cutting-edge technology. Several speakers from enterprise, government and research came together to discuss how engagement by industry will affect the future developments of Grid technology.

Mário Campolargo, head of the Research Infrastructures Unit in the European Commission, spoke about the importance of Grid adoption by enterprise in helping Europe to become the most dynamic and competitive knowledge-based society in the world. "Industry will benefit from a skilled pool of expertise in the same way that academia does ... and the deployment of Grid infrastructures will benefit from its involvement," he said. "E-science processes can adopt industrial practices, from those working in security for example." He expects to see academia forming partnerships with industry based on these mutual rewards.

Companies such as eBay, who deal with huge amounts of data each year, and major banks, which organize millions of daily transactions, are just a few examples of industries that Campolargo suggests will greatly benefit from Grid technologies.

Marin Walker from HP EMEA noted other benefits to industry, including cost savings of up to 80 percent and increased revenue through improved business processes. Hans-Christian Hoppe, representing Intel, agreed: "By adopting Grid technologies, businesses will be able to rapidly adapt to changing usage patterns and form efficient business to business collaborations."

All speakers highlighted the importance of security in this adoption, an important factor for fast-paced industry practitioners who have concerns about job failure and data protection challenges. They will need "guaranteed delivery of data, federated trust and policy management" said Hoppe. With this, the adoption of Grid technology by enterprise will continue to grow and be of considerable benefit to both industrial and academic organizations alike.

---

A Demonstration of Diversity in the Grid

Groups from fields as diverse as biomedicine, earth sciences and high-energy physics are now using the distributed computing infrastructure of the Enabling Grids for E-sciencE (EGEE) project, showing the wide adoption and versatility of Grid technology. This was announced at the EGEE’06 conference in Geneva, Switzerland, along with the prize for the two best demonstrations of applications on the grid.

Of particular interest, and joint winner of the €1,000 prize for best demo, was a monitoring system designed to improve Grid reliability, developed by Julia Andreeva and colleagues at CERN. Their goal: to reduce the number of job failures by understanding and fixing error messages.

Their starting point was Dashboard, a program used to collect monitoring information from sources such as MonALISA, RGMA and GridPP.

"Our system highlights common error messages, which we try to fix to make data transfer much more stable," said Pablo Saiz, presenting the team’s work. The team also create automatic reports showing the performance of each site on the grid. "This can highlight which sites are not behaving properly, so that the problems can be attended to and jobs are more successful," he said.

The second prize-winning poster was the "Climate Data Analysis" project, presented by Kerstin Ronneberger and her colleagues from DKRZ, a service center for climate researchers that offers computer resources and technical support to climate researchers all over Germany. "This project clearly demonstrates how the grid can speed up analysis of data and modeling," said Professor Alan Blatecky, deputy director of RENCI and head of the selection committee for best demonstration.

Another demonstration, notable for its direct effect on public health, and an excellent example of the wide diversity of Grid applications, was a project developed by institutions in Italy and Cuba to help doctors combat Alzheimer’s Disease, the leading cause of dementia in the elderly.

Comparing brain images from suspect Alzheimer's patients with a database of healthy brain images can allow early diagnosis of the disease. Unfortunately, healthy brain images are rare and protected by privacy rules, meaning they cannot be shared. As imaging healthy patients is an expensive procedure, only large institutions can afford to carry out comparative analyses. The team has developed a Statistical Parametric Mapping Algorithm (SMP), "which allows us to use the grid to compare suspect [Alzheimer's] images with data from normal images, collected by a NEST-DD project, funded by the European Commission," said Ivan Porro from the Università di Genova, presenting the project.

This technology not only allows analysis to take place quickly over distributed computational resources, it also enables doctors from smaller communities to access healthy brain databases without moving images between hospitals, thus complying to privacy regulations.

The demonstrations shown at EGEE'06 represent just a few examples of the increasing number of groups using the processing and data storage capabilities offered by the EGEE grid.