September 13, 2004

A one-stop shop for Open Access journals

There are an increasing number of Open Access journals. But tracking down all the ones that interest you on the Internet can be time-consuming. The Directory of Open Access Journals (DOAJ) was created to provide a comprehensive database of Open Access scientific and scholarly journals. The DOAJ project has now entered a new phase and aims to provide article-level records for articles in Open Access journals¹. Open Access Now talked to Lars Bjørnshauge about the DOAJ initiative.

The University of Lund, Sweden, is Scandinavia's largest institution for education and research. The university caters to more than 38,000 students spread across eight faculties and several research centers. Lars Bjørnshauge is the Director of Libraries at the University of Lund and is responsible for all the library and information services at the university. Before he took up the Director position at Lund three years ago, Bjørnshauge worked for many years in academic libraries in his native Denmark. "I was very involved in the early days of developing digital library services," he recalls. "For instance, I was one of the first library directors to cancel print versions of scientific journals in large numbers. I was very much involved in re-engineering academic library services and I closely followed the developments in scholarly communication and new electronic technologies."

Bjørnshauge was drawn to Lund because of the university's strong tradition of developing new services for academic libraries based on new technologies. "We have a number of talented software engineers and very highly-skilled IT and librarian teams. There was a tradition of working on different projects, most of them externally funded by national or Nordic funding agencies or the EU." The university's strong background in information services combined with Bjørnshauge's understanding of scholarly communication led to the birth of the DOAJ project.

During the conference the idea emerged that the scientific community needed a resource where readers could look up freely available scientific content. "We immediately grabbed the idea and said that we would be willing to investigate how this could be set up." Lund seemed the ideal home for the directory as Bjørnshauge headed a team of programmers and IT-skilled librarians. "So, we said - let's set this up. This could be fun. This would be valuable for the scientific community."

Bjørnshauge's team decided that they needed to start by compiling a simple list of Open Access journals. They chose a strict selection criterion for inclusion in the directory. "The journal selection is strict so as to include only truly scholarly content, based on peer-review and the type of quality criteria that are demanded by the Open Access community," explains Bjørnshauge. The DOAJ adopted a definition of Open Access similar to that of the Budapest Open Access Initiative (BOAI) or the Bethesda Principles (see OAN summary). This includes the right of users to read, download, copy, distribute, print, search, or link to the full text of these articles. The DOAJ only accepts journals that offer Open Access to their content without delay from the day of publication. The journal must exercise quality-control over submitted papers through an editor, editorial board and/or a peer-review system, and all articles published by the journal must be Open Access. Bjørnshauge agrees that these strict criteria limited the initial number of journals. But he insists on the need for selecting only truly Open Access publications.

The project began by building a relatively simple database of Open Access journals. "When we launched the service in May 2003 we had around 300 journals," says Bjørnshauge. "It was basically a list of journals with links to the publishers' websites and a classification of the journals into subject categories. It was a rather basic type of service." The database now lists more than 1,200 Open Access journals and Bjørnshauge says it is still not yet fully comprehensive. "We continually get suggestions from all over the world asking us to consider new Open Access journals for the directory. We have a couple of librarians involved in evaluating these journals according to the selection criteria." The directory covers all subjects and disciplines and is organized around 16 level-one subject categories, (for example, biology and life sciences, law and political science, health sciences, business and economics). Many of the journals are in the life sciences and medicine. "But we have all different things," says Bjørnshauge. "For example, in business we have ten journals and in economics we have sixteen."

Bjørnshauge was confident about embarking on Phase 2 and the development of an enhanced service because his department has a lot of experience in working with metadata and importing, reformatting and converting article-level data. "We have another service at Lund called the Electronic Library Information Navigator (ELIN@). We import article-level records from publishers and have created a sort of federated search service where you can search across all the different publishers based on subjects or authors or whatever." The aim of ELIN@ was to tackle the problem of integration of electronic information resources, increase the use of these resources and make the administration of e-resources more efficient. The ELIN@ service was developed for users of the Lund University Libraries and is not restricted to Open Access resources, as of today 10 Swedish Universities are using the service as well. "But we felt that the technology and the service could be valuable for the Open Access community," says Bjørnshauge. The second phase of the DOAJ project aims to integrate article-level data from Open Access journals into a searchable database. When it was launched two months ago the service contained around 45,000 article-level records - it has now grown to over 60,000.

Bjørnshauge's team faced many challenges as they developed the extended service. "Maybe to some extent we underestimated how difficult it would be,' he admits. "We have 1,200 journals and that probably represents about 800 publishers. A lot of Open Access journals are run by university departments or non-commercial publishers. The largest Open Access publisher is BioMed Central with over a hundred journals, but we also we have one journal from a department of the University of Warsaw. We try to help these smaller publishers to mature in their services, so that they can also provide us with article-level data. It's a very labor-intensive process because we have to spend time contacting the publishers and offering advice. We have to be able to receive data in just about any kind of format and convert it to a standard form." The DOAJ database does not host the content itself but provides basic information and metadata including the title, authors' names, abstract, key words, and so on.

"It's fairly advanced, but of course it cannot be compared with subject-specific databases like Medline," says Bjørnshauge. "Today, around 300 journals (25% of all the journals in the directory) have article-level records. And we hope to improve that significantly in the coming months. Our basic aim is to help all those smaller publishers and to train and assist them to operate like BioMed Central does. And this is a long process."

The DOAJ data are open-archive compliant and can be exported to different content providers. Bjørnshauge believes that the DOAJ can help smaller publishers in the long run by making them more visible. "We help them to catch the eyes of the readers, which will attract more citations, and this in turn will attract more researchers. It's a positive spiral."

"Following the launch of the article-level search we have seen a large increase in the number of users," he notes. In July the service was accessed by 40,000 distinct hosts from around 140 different countries. "We can see that libraries and information service providers around the globe are harvesting the data. Just to give you an idea: when we launched phase 2, we performed a search on Google for DOAJ and got 40,000 hits. Now you get more than 200,000." Many of these hits are links to academic library catalogues who have integrated the DOAJ database into their own catalogues. Bjørnshauge says that librarians find this particularly valuable as they can find all Open Access journals in one place.

The DOAJ was initially supported by funds from the OSI and contributions from the Scholarly Publishing and Academic Resources Coalition (SPARC; see Open Access Now, 25 August 2003) and BIBSAM (the Royal Library of Sweden). The service is currently not funded by a particular outside source but Bjørnshauge is actively seeking funding and there are plans to introduce a sponsorship program.

Bjørnshauge acknowledges that Open Access archives and institutional repositories offer additional critical information resources. "So far we have focused on the Open Access journals, but we are in discussion with the open archive community about how a service like ours could be implemented in this area as well. And we hope to come up with some specific ideas," he adds. "There is a need for similar kind of listing of open archives with strict selection criteria and we are working on that. Currently there are a number of lists, but they are not complete and lack a common starting point. We are discussing with several partners about the need for similar service - a directory of open archives. In the ELIN@ service we integrate Open Access journals and archives and information from different sources in the same search. So they definitely can be, and should be, integrated."

"We still have new Open Access journals to integrate and there is a lot of work to be done on the article-level content," notes Bjørnshauge. "Another issue that we will have to deal with now is the so called 'hybrid journals' (which offer an Open Access option to authors who wish to pay for it). We will probably include article-level records for those articles too." He also feels that the DOAJ service could play an important role in archiving content from journals that may have a fragile future. Bjørnshauge and his team are likely to be busy in the months ahead and will continue to develop services that help the Open Access community

¹ Open access publications could be misinterpreted to cover documents, i.e. documents in pre- and e-print archives.

www.doaj.org