Electronic Publishing Software for Libraries

This concurrent session covered the background, purpose, and evolution of the DPubS (Digital Publishing Systems) open source software project, based at Cornell University Library, as well as a case study based on Pennsylvania State University Libraries’ use of the package. The audience left with an appreciation of the potential of electronic publishing software to allow an academic library to provide enhanced services to its user community.

David Ruddy, from Cornell University Library’s Electronic Publishing Initiatives division, who has been involved with the project for a number of years, started by saying that the project had two main objectives:

  • to allow publishers to organise and deliver both open access and subscription controlled content; and
  • to give users the ability to navigate and access content.

The project came about because of a number of factors, including the disaggregation of the publishing industry (mainstream publishers often contract out their electronic services) and the rise in prices of conventional books and journals in the past 10 years. Other reasons for the project included a desire to offer publishing alternatives and offer a tool to allow others to become involved in scholarly publishing, and to support local initiatives in scholarly publishing.

Cornell University Library’s involvement in electronic publishing began as a result of Project Euclid, which started in 2000 and currently provides access to 45 journal titles in mathematics and statistics, involving 30 different publishers. Approximately 66% of the content is available as open access, and the remainder is subscription controlled. The Center for Innovative Publishing is involved with publishing several other titles, and has a number of other projects underway, primarily local initiatives.

The DPubS software began as a project in the Computer Science department, and was picked up the the Library in the late 1990s. Project Euclid provided the momentum for initial redevelopment of the software, and a Mellon grant (in combination with the involvement of Pennsylvania State University) funded further work in the last two years.

There was an ambitious development agenda. The main areas of work were the generalization of the software to support multiple document types, the development of different administrative interfaces, provision of interoperability with institutional repository software, and provision of editorial management facilities. In the first six months of 2006, 6 additional development partners were involved.

The system uses a simple object model, and the architecture is a distributed services model with clearly defined APIs. It supports different presentation options, which can be customized for each publication. It also enables content to be made available under multiple subscription and revenue models (not just open access). It is designed to have low maintenance and operating costs. The integration with an institutional repository means that the repository software can look after preservation and archiving, while DPubs focusses on presentation and access controls.

DPubS 2.0 was released in October 2006 on Sourceforge; it supports OAI 2.0, and can be used in combination with Fedora as an underlying repository.

Further plans include extending the editorial tools to support peer review, enabling it to work with dSpace as well as Fedora, enhancing the administration interfaces, documentation, and allowing contributions from the user community using the open source development model.

The challenges include finding others who are both interested in being involved and have the capability to contribute. David noted that it is a leap for libraries to move from being content consumers to being actively involved in content publishing.

See: http://dpubs.org/, http://dpubs.org/wiki/ and http://cip.cornell.edu for more information.

Mike Furlough, from Pennsylvania State University Libraries, then gave a user’s perspective on DPubS. Penn State has been a DPubS development partner, and their involvement has included testing alpha versions of the software, testing its integration with Fedora and dSpace, developing test cases for journal backfiles and conference proceedings, and refining and testing the editorial services.

At Penn State, the University Press is part of the University Libraries. The Office of Digital Scholarly Publishing wants to provide a scholar driven service, particularly for at risk literature. They hope to experiment with different business models; currently all of their content is available as open access, except for a print on demand facility. Their current project is Pennsylvania History, which has content available back to 1934. They expect to start publishing three other titles in 2007/2008. They are exploring ideas for publishing defunct journals, and setting up a conference publishing service, mainly for conferences hosted at Penn State. They will also consider new original content.

Outstanding questions they will be considering as they use DPubS are:

  • Does the content management architecture align with their mission?
  • How will their implementation contribute to the DPubS community?
  • What staffing levels are needed?
  • How can the publishing program be grown to support the teaching mission of the university?

See: http://dpubs.libraries.psu.edu and http://www.libraries.psu.edu/digital/ scholarlycomm/ for more information.

Q: Is there a space limitation on the number of journals or the number of objects?
A: No. The intent is to keep growing and scale the software up as necessary.

Q: Can DPubS handle journal articles supported by rich data sets?
A: Yes, they have already had this feature requested, and it can be handled by creating another format in the repository.

Q: What is involved in keeping these publishing initiatives going?
A: The journal needs a committed editor and board; it is the responsibility of the content creating community. Preservation and affordability are still works in progress. As journals grow production will require more staff. Library expertise in routine work can help, and it’s a good fit — for example, technical services staff can prepare metadata for journal content. It provides a good opportunity for the library.

Q. What is the role of the subject librarian?
A: Project Euclid works closely with Cornell’s math librarian. Subject librarians can provide advice in particular fields and help with relationship development.

Electronic publishing projects are an ongoing challenge, and require a different mindset. Each project needs to be evaluated.

3 thoughts on “Electronic Publishing Software for Libraries”

Comments are closed.