Academia.edu raises funds to build a Facebook for scientists

Academia.edu, an Internet startup trying to bring the social dynamics of LinkedIn and Facebook to researchers' professional communications, has raised $11.1 million in a second round of funding.

The company's mission -- to overhaul not just how researchers publish results but also how they judge the quality of one anothers' work -- won over Khosla Ventures for the Series B funding. Khosla led the round, Academia.edu said Thursday, and earlier investors Spark Capital and True Ventures also participated. Ben Ling of Khosla will join Academia's board of directors, and Khosla founder Vinod Khosla will become an adviser.

"The goal is to have every single science PDF ever written available for free on the Internet and to build a network of scientists interacting with those papers that will change the face of peer review," the San Francisco-based company said of its mission.

Academia.edu has published 1.6 million papers since its launch in 2008.

Today's publishing system typically involves submitting a paper to a journal that finds other researchers to review it and chooses whether it merits publication. The system has worked for decades, but many have complained that publishers charge exorbitant fees for access to research that's often publicly funded. Those in academia often have subscriptions that grant broad access to many journals, but those outside the ivory tower often have to pay for each paper they want to read. Academia.edu wants to make the process faster and more open.

So where does the money come in? Chief Executive and founder Richard Price expects that Academia.edu's data on what researchers are doing will prove profitable.

"When we've built what we hope will become this new digital infrastructure around scientific research, there are several ways to monetize [it] that stem from mining the vast array of data we will have access to, such as providing information to pharmaceutical companies about trending areas or breakthroughs in science that haven't yet hit their radar," he said. "Another example is that we'll be able to help universities and R&D companies source the top scientists in specific areas, and hence monetize via recruiting avenues."

Academia.edu's membership of 4.3 million researchers are uploading about 150,000 articles per month, which the company said is close to the 160,000 articles that traditional journals publish every month.

Researchers use the site to follow one another's work, track their influence with analytics tools, and, the company suggests, "build powerful brands online."

That brand is what Academia.edu proposes as a replacement for peer review. Instead of having papers vetted through a formal mechanism, Academia.edu uses something equivalent to Facebook popularity. It's a pretty radical departure for an academic culture not known for rapid changes.

Here's how Price describes it:

Peer review is ultimately there to drive a scientist's reputation. Academia.edu's readership metrics provide another arrow in the quiver of a scientist's reputation -- alongside citation counts, and the journal titles they have published under. Over time, Academia.edu will introduce additional peer-review metrics such as comments from other scientists.

We believe that how a scientist's audience interacts with their research online will ultimately be the foundation of that scientist's reputation. A recent feature that Academia.edu added is a way for a scientist to recommend a paper to their followers, and we believe these recommendation counts will be another aspect of a scientist's reputation, something for a grant and tenure committee to look at.

Already, Price said, some researchers believe reputation scores are useful on applications for research grants and tenured positions at universities.

Moving publishing to the Internet also makes papers more fluid.

"We believe that the norms around scientific sharing will evolve in a similar way to that of blogging," Price said. "If a large update to a post is made, readers' attention is drawn to the fact that there has been an update. If the update is small, say just a typo, the change can be made seamlessly."

Researchers share only "pre-prints" on Academia.edu -- papers that haven't been through a journal's peer-review process that precedes publication.

In that way, it's similar to another effort called the LANL Preprint Archive, which launched in 1991 at Los Alamos National Laboratory. That effort grew into Arxiv.org, which is owned and operated by Cornell University.

"Arxiv.org is part of the movement towards building a faster and more open science. It has a lot of traction in physics and math," Price said. "Because Academia.edu's scope is larger, Academia.edu has substantially more papers, users, and traffic."

One way or another, companies like Elsevier that publish scientific journals know change is coming, Price said.

"The scientific journal industry knows that the current model won't last forever. Many of them are innovating by trying new business models, such as taking down the paywalls and charging authors instead," Price said. "I think there is a growing understanding in the industry that radical innovation is table stakes."