March 1, 2016
Is it possible to imagine a social platform for researchers providing a good recommendation system that would also have Impact Factor?
Some time ago I read a piece where the author asked a question, “if academic publishing is about providing services, where is the Uber of academic publishing?” For the author the obvious answer was that there is no Uber-like company in the research communication industry, so it is not about the service there. I do not remember where I found this sentence, but it stayed in my mind, mostly because I do not agree with its assumptions and also because I think it touches on some very interesting problems with the publishing industry. I think that academic publishing is mostly about providing services, and some Uber-like solutions do work in the industry at the moment. But the real Uber of academic publishing is yet to come.
Therefore this piece is going to be another part of “Uberization of everything” and “Uber of X” series, but I hope it will go a little bit beyond that and towards some unexplored possibilities for publishing in the digital era.
In case you have just come from outer space, I will explain what Uber is. Uber is an on-line platform that uses live time data, harvested with mobile software, to connect drivers that have empty seats in their cars with people in need of a lift. It works so efficiently that it has already triggered protests by taxi drivers who feel threatened by the competition from unprofessional drivers. This is quite surprising when we realize that Uber does not actually do that much. As Garry Hall put it, Uber is “selling cheap and easy access to assets that are underutilised”. These assets (free seats and drivers wanting to sell them) existed before Uber entered the market and Uber neither created them or even bought them. Providing a tool to easily discover these assets is the only thing that Uber does.
Publishing as having a lift?
Academic publishing is a little bit more of a complicated system than transport. While transport responds to the need of being transported, academic publishing has several functions and there are a lot of discussions as to which one is the most important. But one can view the publishing market as an environment consisting of two sides: authors and readers. Authors want to be read by (proper) readers, who want to read (adequate) works. The only problem is how to deliver an adequate work to an appropriate reader in an easy to access and useful form. This is one of the major functions of academic journals and book series managed by academic publishers. Looking at this system from the perspective of Uber, one may think that publishers are the “unnecessary middle men”, just like taxi driver corporations. Even if they disappear, authors will still write works to be read, and readers will still be wanting to read them. One may think that what might replace all publishing companies is a perfect algorithm that recognizes the work only when it is written (or uploaded to the system, e.g. from a mobile phone) and bring it to every user that is interested in this kind of work, but only to those specific users, so that readers can always be sure that they will not waste time reading inappropriate papers. These kind of businesses would need a huge amount of data about users’ choices and preferences to operate. In other words, it needs to be a social network which intensively mines data from the profiles of its users and from their mutual interactions.
Well, actually we already have several platforms of this kind in the field of research communication. They are still fresh things, but will they be able to replace traditional publishers one day? Or will academic publishers simply create better platforms of this kind?
The matter of prestige
Regardless of how large an impact on the real dissemination of papers academic social network websites could obtain in future, under the current rules of the game they are unable to takeover the position of academic publishers. The reason is that academic publishers also provide quality control services, and as a result, they brand work with quality stamps.
I often hear people who are neither research professionals nor publishers, using the fact that something was “published in a scientific journal” as a proof of its importance in a discussion. Those who are more familiar with academic code of conduct also use the name of the journal as a part of their argument.
More importantly, grant funders and tenure committees tend to do the same. And despite the fact that qualitative criteria are still part of the evaluation performed by almost every academic body, algorithms and indicators of various kinds have a growing impact on every decision regarding research funding or academic promotion. And of course, these indicators tend to concentrate on Impact Factor and academic journals that might be compared using this metric.
But maybe it is possible to imagine a social platform for researchers, providing a good recommendation system that can also have Impact Factor? Or which is enabled into the academic prestige distribution system in the other way?
Peer review? It is easy!
For sure, to be recognized as a credible place for research publication, a platform of this kind would have to solve the peer review problem somehow. Despite the fact that Arxiv.org has operated without a peer-review system for more than 20 years (and a lot of papers in Arxiv are formally cited before they are formally published in any journal and so before they are reviewed), peer review seems to be a part of the game that is here to stay. The role of Arxiv in modern research communication causes controversies from time to time, such us in case of the recent “Planet X discovery”.
Here’s the danger of releasing un-peer-reviewed papers on arXiv. There are so many reasons why this is likely wrong. https://t.co/aywfgVxIq5
— Mike Brown (@plutokiller) December 10, 2015
But finding a reviewer for a paper is an easy task if you have a database with data about millions of researchers, covering their contact information, research interests, professional and social behavior. So it is possible that one day it will not be a problem for academic social network websites to offer peer review to authors, and become a de facto ‘terajournal’, that might be treated as a legitimate publication venue.
What will happen next?
Have you noticed that we have only one popular search engine in Europe? And we have also one popular microblogging platform? The most popular video website has some competitors, but having a small market share only. And even Google Plus, backed by a natural monopolist in the European web search market, was not able to start a real competition with Facebook.
In the case of Facebook and Twitter, this is mostly the result of a networking effect. One will not just join a social platform X, while his friends have been already using a platform Y for the same purpose. When your friends are on Facebook, you will feel a pressure to join them. When nobody who knows you use Twitter, it will be much harder for you to find a good reason to start with this tool.
But this is also an effect of scale, a huge part of which is the result of the ownership of data (the case of Google). When you have a lot of data you can use it to improve your services. This is why Google search is so efficient. Google knows us well. Their competitors with a smaller user base know us less (or do not know us at all), so it is very hard for them to be competitive.
Both of these effects will take place in the case of academic social networks when they become academic publishers. There is a huge probability that we are currently shifting towards a natural monopoly in academic publishing, a monopoly which will be very different form the traditional publishing companies that we already know.