JOURNAL CROSS-LINKING

The Web's potential untangled

There is a new phrase buzzing around the world of scientific publishing at the moment – the ‘digital research environment’. This may not sound like anything special to the average scientist, who has, after all, being conducting research by digital means, on and offline, for some years. But the coining of the phrase signals proper recognition of that very fact on the part of information providers. And defining such an entity, and therefore a market for it, has also freed them to invest in its optimisation. The upshot has been a flurry of behind-the-scenes activity geared towards enhancing this digital research environment, the implications of which are only just coming to light.

From the scientists’ point of view, the results range from mildly pleasurable to almost imperceptible – for example, a link from some Web resource to something else, exactly where you would have always hoped to find it (where it’s well implemented) or no change at all (where it isn’t). But the new lie of the land can be almost unfathomable from the information specialists’ point of view. The reason is that this enhancement and optimisation has entailed collaborations, alliances, linking partnerships and all manner of other deals, both technological and commercial, between and across organisations. Publishers, producers, distributors and aggregators of online research information have created an interlinked, multi-woven Web of connections. It may well ultimately allow seamless gliding from resource to resource for the researcher – but in the meantime it has also rendered the publishing landscape seemingly unrecognisable.

What we are witnessing is a transition between the old way of publishing and the new. In place of pure competition, there is now ‘co-opetition’ (another buzzword – but it reflects the new mood of nervous détente as well as any). Online journal publishers are beginning to allow users to click through to cited papers from other publishers. They are also passing their material to portals and aggregators and allowing them to facilitate single-article downloads on a pay-per-view basis and to take a nice margin for it. Alliances between publishers and database producers are enabling researchers to download both the articles and the data behind them at the same point. These are just some examples of the moves afoot.

John Wiley & Sons is a case-in-point. Its online content service, Wiley Interscience, has been around for more than four years: offering its journal content to subscribers online is not a big deal. But the last six months have seen new shifts. It announced an agreement with Ebsco, a provider of online subscription management services to more than 260,000 periodical titles, so Wiley’s content will now also be available through the Ebsco gateway. It formed a partnership with LabBook, a life sciences informatics company, to deliver an ‘end-to-end solution’ to life science researchers. This will enable content from Wiley Interscience (including the ‘Current Protocol’ lab handbooks) to be accessed through the LabBook software. And it made an agreement with Celera Genomics to link articles (or references or abstracts) on Wiley Interscience with the Celera Discovery System – meaning that researchers will be able to move from genetic content on Interscience to actual gene data and back again. Wiley clearly wants to be more than a publisher to consumers of life science information.

Eric Swanson, the company’s senior vice-president of STM publishing, outlined the company’s thinking: ‘Firstly, we are very energetically licensing access to our online content to our existing global subscriber base. We are moving from a print production and distribution model to an online distribution and maintenance model. It is having a measurable positive effect on revenue and has occasioned dramatic increases in our investment.

‘This means we can extend our range of customers and usage methods. We want to ensure that there are as many links as possible into our content from other places that users go.’ Herein lies the motivation to collaborate: more exposure in the online world means more potential customers.

Most major publishers of science material are pursuing the same sort of strategy – Elsevier being another example. Elsevier’s online offering is ScienceDirect. This gives access not only to all of Elsevier’s science journal collection (which at more than 1,200 titles is the largest of any single publisher) but, through a variety of agreements and partnerships, also to journals from other organisations, and to a number of databases such as Inspec, Biosis Previews and Beilstein Abstracts. More recent announcements include its science-specific search engine, scirus.com and a new Web-based bibliographic database called Biotechnobase.com.

Amanda Spiteri, ScienceDirect’s marketing director, described the company’s strategy: ‘Previously we were caught up in a onestop shop philosophy, in direct competition with others. That is no longer a sensible goal. Our ambition now is to be the preferred entry point for scientific information and ScienceDirect is key to that philosophy. ScienceDirect has exploded access to our content. The next stage is to continue developing our abstracting and indexing services and getting good disclosure of full text information – both ours and others – whether people are coming in via ScienceDirect or from some other entry point such as an OpAc [Open Access] point in an academic library.’

Of course, in the hard-copy world, one of the most important pointers or ‘entry points’ to any article is another, earlier article – citing it. This brings us to the most significant collaboration among science publishers of them all – CrossRef. CrossRef, which has been going for more than a year, calls itself a collaborative online reference linking service – a sort of digital switchboard that enables connections to be made between papers published online. Reference linking is probably the most impressive example of the way that the new, digital research environment is taking shape, both technically and in business terms. CrossRef comprises a collaboration of nearly 100 publishers who have agreed to put a piece of digital information relating to the position of their online papers on the CrossRef database. This information is termed a Digital Object Identifier (DOI) and enables a user to click from an article reference in one paper, held by one publisher, to the referred article (which may have been published elsewhere) or at least some information on how to obtain it, if the user does not subscribe to the journal holding it.

Amy Brand, CrossRef’s business development manager, explained: ‘Citation linking enables the user to navigate the scholarly record, moving from one article or idea to another across journals and publishers. It can be a matter of a click or two to get to the full text.’

It sounds just like navigating links on the Web, but it’s not quite the same. ‘A key benefit is that a DOI is a persistent link, unlike a URL,’ continued Brand. ‘The CrossRef database holds the threepart record for a publication, consisting of metadata, DOI and URL. If the URL changes, the publisher simply informs CrossRef, but the DOI link that has already populated the Web remains constant. This system uniquely avoids stale links in citations or database records.’

Both Elsevier and Wiley are prominent members of CrossRef. In fact Eric Swanson is CrossRef’s chairman. ‘CrossRef means that publishers make just one agreement and link outwards indefinitely – a sort of hub-and-spoke approach,’ he said. ‘The idea is to simplify channels and effect standardisation.’ Elsevier’s Spiteri echoed his enthusiasm: ‘We know a lot of incoming traffic is generated by the CrossRef gateway. We are loyal to CrossRef because people want to see industry-wide protocols and this is what CrossRef offers.’

CrossRef is run on a not-for-profit basis, and survives by the fees from its member publishers and DOI registration and look-up fees – also costs borne by the publishers. So publishers are actually paying to allow people to click from their material onto that of other publishers – of course, in the hope that they get more clicks in than out. For a company like Elsevier this is almost guaranteed – of the 3m links within CrossRef, 1.7m of them refer to Elsevier material. ‘We’re bound to be a big beneficiary of this and also a major financial supporter of the initiative,’ confirmed Spiteri. But for others ‘co-opetition’ can entail more risk. Reconciling the need for cooperation with a competitor, says Swanson, can be something of a conundrum. ‘We are constantly wrestling with where the border between competitive and pre-competitive is. Too competitive and we will be doing our readers and authors a disservice. Too co-operative and we get thwacked by our legal advisors. But we like to think that a rising tide lifts all the boats. We deal with the issue delicately but with some brusqueness and acceptance of risk.’

Regardless of which companies thrive in this new atmosphere, users will undoubtedly benefit from the initiatives. For a start, online subscriptions mean a huge increase in the number of people in an organisation that can have simultaneous access to the content, compared with, for example, the process of queuing to photocopy pages from a single copy of a journal. And then, providing their organisation holds the appropriate journal subscriptions, researchers can move effortlessly between citations. Meanwhile, libraries can adjust their collections according to users’ needs properly for the first time, because usage can be accurately monitored. In addition, the new method of distribution often brings with it a new pricing structure.

Elsevier, for one, used ScienceDirect as a good excuse to talk directly with its customers and throw out the restrictive pricing structures and agency deals on which the industry had previously been based. It resulted in better value for money for the customer but a rather raw deal for the agents, admitted Spiteri. So far, so good. But there is more. CrossRef is not the only gateway and publishers are not the only purveyors of information. Reference linking by a different technology protocol is also provided by PubMed Central, a service of the US National Library of Medicine, which provides access to more than 11m Medline citations and some from other life science journals.

PubMed also links in with a number of genomic and protein databases from the US National Center for Biotechnology Information. Meanwhile, there is ChemPort from Chemical Abstract Service (CAS) in the US, another old hand at reference linking, this time for chemistry content within the 2,200 ChemPort member journals, and again, using different protocols. This also has the advantage of linking via CAS registry numbers and the CAS bibliographic database. Admittedly, these are subject-specific gateways, but there are significant overlaps between their members and those of CrossRef.

So, have the arrival of CrossRef and the myriad of other collaborative initiatives complicated things for CAS and its ChemPort service? ‘No. It’s more interesting, though!’ admitted Eileen Shanbrom, CAS’ Web content manager. ‘ChemPort predates CrossRef and actually a lot of other thinking on linking standards. But we will write to other people’s standards if our users want that. Our mission is to compete in the digital research environment. New alliances and initiatives bring more opportunities because they bring more users.

‘We already have a strong collaboration with publishers but now that collaboration is different. We are dealing with different people within those organisations – new business development people, online enhancement people, and so on. Sometimes they need educating about who we are but the main improvement is that then we can point back to their literature electronically. ChemPort acts between publishers and users – it is not competitive.’

But is CrossRef a competitor? ‘No – we have joined CrossRef.’ Confused? Shanbrom admits that lots of people are. ‘But we are working closely with information professionals and are educating one another,’ she added.

‘Scientists themselves are only just beginning to see the bene- fits of reference linking but it will simply make things easier for them. It’s the information professionals and librarians that are having to work hard to understand the options.’

As if this isn’t enough to assimilate, we have yet to touch on portals and aggregators. Take Ingenta, a rapidly growing, wellfunded UK-based organisation that hosts journal content from hundreds of publishers and distributes it to users on a pay-per-view or subscription basis – a sort of journal portal.

Ingenta presents yet another route to research information, and, unsurprisingly, also likes to portray itself as the place everyone should start looking. It is also a member of CrossRef. Different again are aggregators, such as Ovid – itself a subsidiary of a major scientific publisher, Wolters-Kluwer. Ovid’s holy grail is the licence to host and distribute other people’s content of all descriptions – as long as it’s good enough. ‘We look for premium content – databases, electronic journals, and e-books,’ explained Andrew Richardson, general manager of Ovid.

‘We tend to load up the full text in SGML onto our databases. Sometimes we do link to the publisher but the idea is that users stay within Ovid. This makes navigation between and within journals easy. We take the view that if they leave Ovid to go to a publisher site, their research journey is over.’ This couldn’t be further from the views held by the protagonists of interlinking – and yet Wolters-Kluwer is a member of CrossRef.

‘It’s true that it’s a complicated environment,’ said Richardson. ‘We must compete and provide in the same breath. Ovid’s strengths lie in the quality of our content and the richness and utility of the user experience. We’re not the cheapest but we’re among the best.’

Clearly there is a lot of settling down to be done before the new categories of scientific information provider become apparent and the bands of allegiance are finalised. Spiteri summarises neatly: ‘There is a fundamental conflict in the market. Customers want more simplified and unified access, but not a monopoly. We will see some sort of consolidation. It’s hard to see where the balance will lie between competition and cooperation and also whether companies will diversify or stay within their niche.’

In the meantime, researchers can sit back and watch the links arrive on their desktop and click, click away. Information professionals will probably already be taking a more active part in the evolution of the business – and those who are not should perhaps be considering what they want out of it.

Standardisation of systems and methods, in a way that suits everyone, must surely be an aspiration. Either way, while information providers are scrambling over one another to be the first to truly meet the needs of their customers, there could never be a better time to influence the market.

back to main features page
home