Getting serious about open access discovery — Is open access getting too big to ignore?

Aaron Tay
Academic librarians and open access
9 min readJun 19, 2017


Note from 2020 : Three years on, a lot of what I wrote in here has come to pass, with pretty much every majority discovery system (including library ones) having made big strides to include open access content. There is still progress to be made in terms of standards and consistency of how these content are included of course. I leave the content below unaltered for historical reasons.

Unpaywall — perhaps the most popular extension for finding full text

There seems to be something in the water. Out of the blue, interest in helping users find free/open access articles seems to have blossomed since earlier this year when Unpaywall was unveiled.

With all the intense interest Unpaywall is getting (See coverage in academic sites like Nature, Science, Chronicle of Higher education, as well as more mainstream tech sites like Techcruch, Gimzo), you might be surprised to know that Unpaywall isn’t in fact the first tool that promises to help users unlock paywalls by finding free versions.

Predecessors like Open Access button (3K users), Lazy Scholar button (7k Users), Google Scholar button (1.2 million users) all existed before Unpaywall (70k users) and are arguably every bit as capable as Unpaywall and yet remained a niche service for years.

And yet Unpaywall is getting breathless headlines like “How a Browser Extension Could Shake Up Academic Publishing” from Chronicle of Higher education.

The Gold rush of services built on/around open access

To be fair, Unpaywall was built around the very cool service, which when given a doi, could give you a link to a free version. It even has a nice API service. But even that wasn’t unqiue as a predecessor service — service existed that did the same thing.

Still for whatever reason, suddenly services built around helping users find free full text began to emerge all at the same time. Canaryhaz a commercial service that helps direct you to articles you have access via your institutional affiliations and open access versions was the next logical step.

Canaryhaz extension — covers both free and paywall versions via your institution

Similarly, lean library (a company founded by former librarians and targeted at libraries) is now offering a browser extension that focuses on making access to articles behind paywalls more seamless via institutional subscriptions. But in a nod to the times, it also helps find free articles when institutional subscriptions fail.

Lean library browser extension

The return of the browser extensions

As a librarian, who came into librarianship almost exactly 10 years ago I find this development pretty amusing. Back in 2007, libraries was in the middle of experimenting with browser extensions (as perhaps part of “Library 2.0” movement)

I remember playing with opensearch plugins, browser toolbars (anyone remember Conduit toolbar? Google toolbar?) and of course the proxy bookmarklet (of which Lean library’s extension is meant to be a clear improvement on). All in the hope of giving quicker access to library services for our users.

Custom toolbar I created in 2009 using Conduit toolbar — ah the memories

Back then I draw inspiration from the great Dutch Medical Librarian, Guus van den Brekel who was in my book leading in the exploration & use of such tools for his users. To this day, you can find my earliest blog posts were on such subjects, but by 2010 interest began to cool.

Browser toolbars were seen to be clunky and worse mostly sources of malware and we turned our attention towards social media and handling mobile. Today the only survivors of those heady days are Libx which itself seems to be barely afloat and the idea of proxy bookmarklet , though in both cases they are not usually well known outside the library community.

Library discovery of open access gets serious

The main drawback of browser based extensions if you ask me is that you could never get even a large minority of your userbase to install and use it.

Hence it is important that your discovery service cover open access and free items.

Sadly, library discovery tools had never been good at this.

I always thought it was a supreme irony that a outsider tool — Google Scholar did discovery of open access better than our own library search tools — in particularly making contents of institutional repositories properly discoverable (only showing full text items).

Over the years, as I got better at understanding Library discovery , I could see many reasons why. Due to quirks of history, our technologies didn’t consistently require full text to be identified making it difficult for aggregators to combine them. (The NISO standard appears to be very new and unsupported).

Even with that overcome, traditionally, our link resolvers and knowledge bases worked at the package or at best journal level and not article level unlike Google Scholar and hence gets fits when trying to deal with hybrid journals with a mix of paywall and open access articles in the same issue or worse yet Green OA articles hosted in institutional repositories.

We just weren’t doing well at what I dubbed Library discovery and the open access challenge.

As any librarian familiar with the situation can tell you, while in theory Discovery services like Summon, Primo allow you to add a lot of open access material from various open access sources, in practice you wouldn’t do so because there were many issues such as inability to reliably identify full text items among non-full text items in open access collections as well as outright broken links. This was particularly bad for institutional repository contents (if they were even included, indexed individually or as part of a aggregator).

Things seems to be changing though. First off, Scopus and Web of Science began to tag articles as open access. Still initially this was done crudely based on whether it was in a Open access Gold journal. But this still missed out hybrid articles and even more seriously Green Open access material.

Similarly Proquest’s Summon one of the 4 major web scale discovery services also recently implemented a Open access filter.

Virginia Tech Libraries — Summon with Open access filter turned on shows 13 million results

More importantly Proquest (which owns the two main library discovery services Primo & Summon) announced their serious intent to improve this aspect of discovery.

They recently started to solicit feedback from Primo and Summon customers on open access discovery in their discovery services with interested users given access to a online (Basecamp based) special interest discussion group.

Besides indexing in discovery services , link resolvers are getting into the act too. oaDOI intergration for SFX resolver exists and it seems the same will be coming for Alma Uresolver as it has been added to development plans.

What how will it actually look like? Recently Proquest updated their 360 Link link resolver to integrate with both oaDOI and the open access button to find free full text.

More exactly, it will query the two systems to check if there is a free full text item and display the buttons only if available. This check only occurs I believe if the link resolver is unable to find suitable access via Library subscriptions.

As hinted above getting discovery of open access to work well given the way our systems are currently designed isn’t easy and the documentation of the open access filter in Summon gives you a taste of the issues.

More services around discovery of open access are emerging

Besides traditional companies in the library field, I’m also looking with interest the rise of companies like 1Science which offers newservices around effective indexing of open access items (Oafinder), automatically fill your Institutional repository with open access items by your authors (Oafinder+), help decide on subscription decisions based on amount of existing open access items out there for the title (OAfigr)

Is Open access getting too big to ignore?

The two developments above are actually quite remarkable. When commercial services start sprouting up or caring about something, I tend to sit up and take notice.

The emergence of such tools that try to find Open Access articles is an implicit recognition that open access articles are in such numbers where they can no longer be ignored and it is worth the effort to install such tools to find them.

Go back 10 years or maybe even as recently as 5 years, most mainstream researchers and many librarians (those not directly in Scholarly communication) would consider discovery of Open access articles as at best a second thought if at all but it seems things have changed. Either there is now such a big pool of open access articles that they can’t be automatically ignored and/or there is much faith and belief that the percent share of open access articles is going to continue to grow.

A virtuous cycle ?

I’ve often noted that there Green Open access (and in particular institutional repositories) tend to faces two issues.

One is the reluctance of researchers to deposit their articles.

Secondly, even if the article existed in Green OA form , there was the difficulty of discovering whether such articles exist as a substitute to the paywall article and many Institutional repository contents would be missed if not for Google Scholar.

I am perhaps trying to be controversial to say that Google Scholar alone has done as much if not more than most librarians or open access advocates to raise awareness of Open access by making the content in Institutional repositories discoverable. Even more controversial would be to credit the likes of Mendeley, ResearchGate, for making sharing/depositing of papers online by researchers a norm.

Both forces together of course with the hard work of open access activists and librarians have perhaps managed to push up Open access to a level where it cannot be ignored and be worth searching when faced with a paywall.

Tools like Unpaywall and its cousin further work to create a virtuous cycle. When more researchers find it easy to get open access/free alternatives , the more likely they will see the point of putting their papers online which will further increase the effectiveness of such tools. Non-academics who do not produce papers, will see better the value of open access.

Implications for libraries

For a number of years, I have wondered about the possible impact of open access on libraries. Assuming open access is inevitable (I would say this is still not 100% settled, particularly the final level open access will stop at ), a very important question to consider is how fast the transition will be.

As a librarian I would prefer a more gradual transition , giving libraries and librarians more time to adjust to our new roles and new environment and to phase out old functions but ideally how and when open access happens shouldn’t be influenced by our selfish desires.

In any case, I’m always on the lookout for possible signs pointing to the speed of transition .

Would the fact that we are now seeing the emergence of commercial services that using discovery of open access articles as a major draw, tip us off that open access will dominate or at least be a big part of the future? Certainly these commercial services are betting on such a future and we would be foolish to discount it.

In my next post, I will consider options librarians have in a mixed, hybrid open access-paywall world . I will argue that, in a world where open access cannot be ignored, a lot of interesting discovery options open up and we librarians will have some interesting decisions to make.



Aaron Tay
Academic librarians and open access

A Librarian from Singapore Management University. Into social media, bibliometrics, library technology and above all libraries.