Science's experiment in publishing

Will a database of academic research with a social networking twist pass the test among scientists seeking publication, asks LISA O'CARROLL

IT’S BEEN billed as the iTunes of scientific research and this week was named one of the top 10 technology companies to watch in Europe by a giant of the dotcom boom.

But Brent Hoberman, founder of lastminute.com and now a venture capitalist, is just one of many big names buzzing about Mendeley, a London-based start-up.

Last month a White House director, Beth Simone Noveck, said it wanted to see this new tool adapted for policymakers in Capitol Hill, forcing them to be even more transparent on the background to their work.

So what’s it all about? Put simply, Mendeley is an open-access searchable database of academic research with a social networking twist. Every paper potentially comes with a review context – what academics read it, what else they read – and functionality to enable scientists to comment and link to other similar research.

It enables scientists to bypass a lengthy process of traditional publication in journals such as Natureor The Lancetand network with peers easily. Anyone can submit their papers to the system – up to 1GB of space is free, with a small premium for greater space and customised rates for universities.

Founded less than two years ago, the service already has 456,000 users with 33 million research papers – comparing, on the face of it, well enough to Reuters, formerly Thomson Scientific, which has about 40 million papers in scientific fields on the web.

There's no doubting the quantity of papers is impressive, but the question is, can Mendeley be of real value to a community that relies on esteemed journals like Nature, The Lancetand Cancer and Cellfor validation?

As Trinity professor John Boland, a director of nano-science project Crann says, the difficulty will be separating the wheat from the chaff.

The German-born founder of Mendeley, Victor Henning, is being funded by the original engineers behind Skype – no slackers when it comes to disruptive technologies – and he’s convinced he is on to something.

Scientists who have used it rave about it and say that as long as publication in Natureetc isn't your Holy Grail, the service is an eye-opener – for a start it wipes weeks of tedium from research processes, particularly in building bibliographies.

Tony Reid, lecturer in nursing and a PhD student at Waterford Institute of Technology, has been using it for six months.

“It is the most user-friendly bibliography tool I have ever come across. You can just start using it without any preparation. It does a lot of things automatically that otherwise you would have to do manually.

“It will go through your documents and go through the citations, extract details on the author, titles, ISBN numbers etc. The time it saves is amazing. Normally it might take five minutes for every pdf and this takes seconds.

Reid is doing a doctorate in clinical psychology and says it could take him “days, weeks or months” to go through the “hundreds and hundreds” of pdf files he is using for his PhD.

“That alone is a wow factor for me, it’s worth its weight in gold,” he says. All you have to do is just drag and drop pdfs into Mendeley Desktop to start the automatic extraction of bibliographic metadata.

Like Facebook, which successfully invited its users to translate from English into dozens of languages, Mendeley is relying on its community to enhance the offering further.

Founder Henning says Mendeley is working on a sample project with a neuroscience department at Harvard to assess all the papers around Alzheimer’s disease.

“Normally you would have to read a paper to find out if another piece of research or data is being cited because it supports a theory or it’s being cited because it’s bad. The team at Harvard are going to create semantic links which will give each citation a meaning – tell you whether a supports b, whether a contradicts b, whether it uses the same methodology and so on,” he says.

Henning is hoping this will mobilise the community to start doing the same with every other paper in the system, applying what he calls “sticky notes” to the system potentially line by line. If it works, it could create a review system that could rattle bosses at the likes of Elsevier Publishing, the leader in scientific and medical literature.

“Their business model is ingenious. They get their content for free from academics who want to be published, they get all their reviews for free because it is such an honour to be invited to review a peer’s work in one of these journals. Then they repackage the information and sell it back to the academics,” he says.

Scientific publishing is one of the most lucrative in the world, with turnover worth up to €15 billion a year. Irish universities spend millions subscribing to these journals every year which are an important part of the academic validation process.

Trinity’s Prof Boland believes it will be difficult to dislodge the scientific journal’s powerful position in the academic world.

“It is important to understand that while research is collaborative, it is also enormously competitive. The real cutting-edge scientists are all searching for the grand prize and they won’t be sharing it,” he said.

Henning agrees, but says the model is shifting and in medicine a system called PLOS1 is now on his wavelength.

“It establishes whether the research is technically good, not whether it has impact. The wisdom of the crowd decides that,” says Henning. That, he believes, is the future.