How 'Google Science' could transform academic publishing

Shutterstock

At least, that is apparently what one individual wants us to believe.

Wired.co.uk is in possession of a document, sent anonymously, detailing how "Google Science" would bring together existing services such as Google Docs, Google Plus, YouTube and more to create a platform that challenges the paid-for model of scientific publishing and provides academics with an opportunity to connect with each other more efficiently. The document was allegedly given to a handful of academics in Berlin this week by Google executives -- so says the email sent to this establishment and a number of other sources.

A name appears in one of the screenshots purporting to exhibit Google Science in action -- Dieter Krachtus -- and Wired.co.uk contacted him to find out if the document is in fact false and mocked up. (There's also a smiley winky face somewhere in the presentation, and a typo, so we were not totally sold...)

Krachtus has since responded to deny sending Wired.co.uk the document, but reveals that the presentation did in fact belong to a 2011 "Google Science project" he prepared for "a couple of friends and acquaintances at Google". The document, is exactly the same -- bar a date change.

A Google spokesperson is currently looking into the validity of a burgeoning "Google Science" project, but so far has been unable to find anything and has no comment. Krachtus believes the whole thing is a prank being played on him. But the email and document appear to have been sent to a great number of journalists and industry players. And still, the origins of its sender, remain a mystery.

It is of note, however, that Head of the Strategy Unit at Research Councils UK Alexandra Saxon, told Wired.co.uk that she has been hearing rumours about such a platform for the past six months to a year. "There have been lots of rumours but nothing substantiated yet. No one that I have been party to comments from has had anything particularly disparaging to say, more waiting to see what will come from it."

Whether the email Wired.co.uk received was the doing of an open-access proponent, a Google employee attempting to stir trouble, an academic who was genuinely presented with the article at a meeting, or a rebel without a cause -- it does raise an interesting debate that has been brewing for many years now. "It is easy to imagine a drastically accelerated scientific progress if we were to fundamentally improve the way science is done on every level," reads the presentation. "By fertilising all fields of science we would pave the way for a myriad of great ideas and new inventions (one would not get otherwise or too late)."

It continues by detailing the increasing interest in collaborative, open-source models, such as Wikipedia, propelled forward by the proliferation of the internet. "The great anecdote of the printing press is that it immediately led to pornographic novels but it took over a hundred years to get to scientific journals. Imagine where science would be today if it were not for this delay! ... and are we today repeating a mistake when it comes to the internet?"

The dramatics are well-founded. Scientific publishing has been ripe for disruption since it was first put behind a pay-wall -- that's a lot of tradition to break. Startups in this space are already working to undo the domination of the paid journal mode, with Figshare most recently launching in the UK to open up the sector.

Digital Science, part of Macmillan Publishers, has its own technology hub where it helps startups, including Figshare, develop software and apps for the scientific community. Some of the biggest players in this space include the journal PLOS, which has published close to 90,000 peer-reviewed open access papers.

On top of this, the highly prestigious Nature Publishing Group says that 38 percent of the research articles it publishes are now open access. But while more and more governments are making open access mandatory for publicly funded research -- Research Council UK launched a new funding procedure this year to help academics fulfil the UK's open access policy -- Elsevier, one of the biggest academic publishers in the world, made $3.2 billion in 2012 from its more than 2,000 journals. As the alleged Google presentation points out, in spite of this "99.9 percent of the work is done by scientists".

This is why Google, one individual wants us to believe, is best placed to step into the breach to "tap into the possibilities of the internet and unleash the power of collective action on science", basically by adapting its own tools and the things it's learnt along the way.

What would "Google Science" look like?

"Huge parts of the current system are harmful to the scientific process and have to be replaced," the presentation claims. So Google would do so, naturally, with its own tools. The proposed platform would integrate Google's search algorithm to make discovery far better, and increase collaboration between scientists by integrating Google Plus. "After establishing a social network of scientists, its collaborative power will produce the tools of the future," the document predicts. "Google Science" would launch a number of journals, be "self-organising" and yet have a team of "qualified reviewers". "99.9 percent of the work, including peer review would be done by the scientific community," the presentation says, before paradoxically adding, "Human interaction rarely needed".

For one leading industry source, this whole description instantly raised a number of red flags. "These platforms are only ten percent tools and technology, and 90 percent community," they told Wired.co.uk. "You really need someone to organise things. Have you ever tried to get support for any of Google's services? I tried and what a nightmare. I spent an hour and a half on the phone. Having a community that is self-governing doesn't gel with my experience."

The proposed service would be invite-only at first, and contributing scientists would need to use real names so there could be no anonymous comments or ratings. There is another rather significant issue here. This contradicts Google's July announcement that Google Plus would finally be doing away with its real name policy for signups -- this would surely bring challenges to integrating the social networking side of things the document talks about. However, Google Scholar obviously uses real names. So it's feasible, if Google did want to launch such a service, there would be an added layer for anyone that wants to join.

The inclusion of real names would be significant, because besides the hypothetical platform's peer review process, the impact factor -- used to evaluate a paper -- would be created by integrating comments, ratings (on things like whether the study is reproducible, etc.) and Google Analytics.

The publishing community is, however, already rather familiar with integrating innovative metrics.

Iain Hrynaszkiewicz, Head of Data and HSS Publishing, Open Research at Nature Publishing Group says: "We provide quite a sophisticated collection of article level metrics, from the number of page views or downloads, to individual article level citations from different sources. Altmetric.com pulls together all kinds of online attention from articles -- the number of tweets, blogs, or Facebook mentions, for instance."

On the addition of a comments section, Timo Hannay, Managing Director of Digital Science -- a Macmillan Science and Education subsidiary that supports startups in the scientific software space -- notes that other publishers have introduced similar mechanisms, with no real impact. "That's not to say there isn't a scientific conversation online," he adds. But aggregating these comments from across the internet, he suggests, is where the value lies. The industry source Wired.co.uk spoke with agrees that those conversations outside of the academic sphere, are perhaps more important. "People post stuff to Twitter, but they tend to want to keep these things separate."

Why would anyone switch to "Google Science", then?

Most major publishers have an open access arm of some sort, and many integrate the aforementioned features already. So why would anyone switch to a tech company, over an established academic journal? Google's reach, and its technology, are obviously benefits. But Hannay suggests far more would be needed to persuade 100 percent of scientists away from the current model, and to another one -- be that one brought to us by Google or another company entirely.

The problem, he says, is not that there are too few options to publish in an open access format. It's that most academics don't think about it too much. "Most [academics] don't particularly care about open access, in part because they are not incentivised to do so. This is changing, but only slowly, and right now most still care more about publishing in established, high-profile journals and in gaining a lot of citations." "Perhaps the biggest problem for academics is the time and effort involved in writing and publishing a paper, particularly the peer review process. Academics don't publish fewer papers because they can't afford to pay for publication."

If Google, or another company, had a secret weapon to disrupt the peer review process, now that would be worth getting excited about. "Peer review is the thing researchers complain about the most -- no one has come up with a way to do it significantly better so far," says Mark Hahnel, founder of Figshare. "This is something they'll have to overcome -- can they improve what the current models are? There's pretty much nothing you can think of that's not been tried by someone at one time."

Stuart Taylor, who heads up the Royal Society's publishing department, says "journal prestige comes from the rigour of the peer review they carry out" and believes that Google would have to "do something really radically different" to make an impact.

Hannay agrees, adding: "It doesn't sound terribly unique compared to what others have already tried, and what's going on elsewhere. But maybe Google can do a better job. "Google doesn't do this for a living, they currently do search for a living... Can they come up with a new model for peer review, with a smarter use of technology and social dynamics? No one has completely cracked this problem yet but maybe Google can. It's certainly good for people to try."

This seems to be the consensus from everyone Wired.co.uk spoke to about the potential platform. That the approach is nothing new, but propelled by Google's technology it could be interesting. If nothing else, it raises an important debate and important questions on how best to push the open access movement forward.

How likely is it that Google penned this presentation?

It's almost certainly a hoax. Krachtus tells us he has had nothing to do with the presentation since 2011/12 when he shared it with friends to get feedback, then Google. But when he originally published it, he did encourage people to change it, re-share it and even remove his name -- so whose hands it ended up in before the mass email went out, is anybody's guess.

In spite of all this, Google's past activities in the space -- including with Google Scholar -- also suggest it's entirely likely the company has an interest in open access publishing.

Back in January of 2008, the tech giant announced that it would begin hosting terabytes of open-source scientific datasets at http://research.google.com, for free. Google Research Datasets -- or Palimpsest, the catchy name it was initially known by -- planned to make use of Google algorithms to help others analyse the data, and commenting and annotating options would be integrated. The system would actually involve carting a 3TB drive array around the world so that anyone who needs to upload vast amounts of data, can. Twelve months after the announcement, however, the programme was publicly shuttered.

There is also the Google Plus post published earlier this year

by Google Research engineer Kayur Patel, which described how he and his fellow Googlers built an "interactive, collaborative analytics tool by integrating Google Docs, Chrome, and IPython".

Integrating IPython Notebook, a web-based, interactive environment where you can combine all kinds of rich media along with text and code, sounds an awful lot like the additional layer Google Science would deliver to tie together all its platforms to create a new one. It fits, that in his post Patel -- who was working alongside Corinna Cortes, Head of Google Research, NY on the project -- says the team had been working for a year "to understand how people collaborate on data analysis and to build better tools to support them". "There is zero setup, because all the computation happens in Chrome. You can even quickly and easily package your analytics pipeline into a GUI for folks that don't want to program. In effect, you can go from zero to analytics with little impedance," he writes.

If this were the early murmurings of a Google Science platform, they would be very early. Patel goes on to say that it will be open source and people can build their own platforms in Chrome. "It's worth noting, that not everything Google does is long lived," says Digital Science's Hannay. "I wonder what level this

[Google Science] idea has reached in the organisation. To their credit, Google encourages all sorts of wild ideas and most of them don't happen soon -- or at all. Think of the Loon project, which was first mooted many years before it launched."

Given that Google's Wi-Fi-broadcasting balloons, as fantastical as they sounded eight years ago, have been airborne for some time for testing purposes, if Google Research Datasets were the precursor to Google Science, we'd be about on track to see something concrete come to fruition.

Google has always been a proponent of an open access world --

Google Art Project, launched in 2011, sought to bring the world's most famous artworks online through partnerships with the likes of the Tate in London, MoMa in New York and the Uffizi in Florence.

The alleged Google Science pitch Wired.co.uk is in possession of refers to a remark made by Larry Page at the American Association for the Advancement of Science conference in 2007, on just how much this issue effects the sciences and scientific progress. He said: "We have to unlock the wealth of scientific knowledge and get it to everyone. I don't care what we do, but we need to do something."

Many, would agree with this. RCUK's Saxon, says: "At the heart of our policy is the aim to make all publicly funded research open access. This will have an impact not only on the science/research itself, but will also provide benefits to businesses that can more quickly take advantage of cutting edge research, and also to society which will be able to access the latest thinking."

Could Google really be the entity to one day facilitate that epic impact?

The Royal Society's Stuart Taylor, who has overseen the launch

this year of Open Science, notes that while it "makes perfect sense given where Google was going,", a potential platform would all depend on "how well received by the scientific community it would be", considering how journals have raised their reputation over the years. "They have the potential to definitely try. This is a time of great flux. We are not simply tied to journals and traditional articles now -- there are pure data publications, people are getting much more used to sharing their datasets; there are nano publications, where it's literally just facts rather than entire studies. "There is room for all sorts of new players. The proof will be in what the scientific community make of it and how valuable they find it. Things are introduced and not taken up. Google is well-placed though."

This article was originally published by WIRED UK