At the recent Berlin 9 Open Access meeting, a pre-conference session on open access publishing featured speakers who detailed the required innovations in publishing business models necessary to both make scholarship freely available and to ensure sustainability. Among the speakers was Dr. Neil M. Thakur of the National Institutes of Health. His presentation centered on an aspect of open access that I have not seen discussed before. Thakur opened with a central question of how to do more with less and he listed three options: work longer, work cheaper or create efficiencies in productivity. It was the latter (and only realistic) option that he concentrated on. Making scientific publishing more efficient requires open access to the literature but for reasons that have previously been overlooked.

In the past, advocates for the open access to scholarly literature have emphasized two audiences which suffer for lack of access to literature: scientists who work at under-funded organizations and who are unable to afford increasingly high subscriptions to scholarly journals, and motivated citizen-scientists (sometimes patients with debilitating diseases) who take it upon themselves to learn the technical language of their area of interest but who are locked out of a large body of literature due to a lack of resources to pay.

But Thakur brings in a third and until now ignored audience: machines. The development of natural-language computer processing and text-mining services is going to be increasingly useful in science in the near future. Because most researchers now face an information-glut rather than an information-scarcity, it is more and more important for them to be able to scan and review large bodies of publications which cannot be covered by simple linear readings. So this time-scarcity problem can be addressed by making the text of scientific publications open to machine processing and interpretation in order to allow scholars to quickly review publications both past and current based on the frequency of certain terms, their proximity to one another and other algorithms. This machine-to-machine access to scholarly literature is a productivity multiplier, Thakur said in his presentation.

A second presentation was by Peter Binfield from the Public Library of Science (PLoS). This is one of the most accomplished open access publishers using the business model where the author pays an article processing charge. In addition to this new way of doing the business of publishing, in recent years a new journal, PLoS One has become the largest journal, publishing over 6000 papers in 2010*. (Binfield expects to publish more than 15,000 in 2011). Despite the high volume, this journal publishes only papers of sound scientific quality and all manuscripts are peer- reviewed as with any other scientific journal. The key difference is that there is no editorial oversight filtering submissions based on popularity or widespread appeal of the subject matter; no matter the topic, if the science is done properly and it passes review by other scientists, it can be published in PLoS One. This model has become so popular that it has spawned a number of imitators from both commercial and non-profit publishers and Binfield pointed out that most of them have article processing charges nearly identical to PLoS One ($1350)

Interestingly, PLoS One was assigned an Impact Factor^® by Thomson Reuters in 2010 and although the Binfield says that PLoS doesn’t particularly care for the Impact Factor^® as a useful measure of scientific achievement, the inclusion of the journal in this popular metric probably explains the spike in submissions during 2011.

*According to Smithsonian Research Online data, Institution scholars have published more than 65 items in PLoS One including 25+ in 2011.