The Open Journal Project
- Project web site
-
http://journals.ecs.soton.ac.uk/
- Programme area
-
Electronic Journals
- Contact details
-
Prof. Wendy Hall,
Department of Electronics and Computer Science, Southampton University, Southampton, SO17 1BJ
Phone: 01703 592388 Fax: 01703 592865
Email: wh@ecs.soton.ac.ukDr Les Carr,
Department of Electronics and Computer Science, Southampton University, Southampton, SO17 1BJ
Phone: 01703 594479 Fax: 01703 592865
Email: lac@ecs.soton.ac.uk
Project description
as of May 24th 1997
Introduction
The reference or citation is an integral feature of the modern academic journal. The unique value of the authored citation in the expanding scholarly literature was recognised over 40 years ago by Eugene Garfield through his development of citation indexing. As publishers begin to make the mass of print journals available over the World Wide Web, the impact of citation indexing is set to be repeated in the form of the citation link, a hypertext link taking the reader from a reference in an online paper directly to the cited paper in the instant of a computer mouse 'click'. When fully realised this feature has the potential to transform journal usage. The Open Journal project is demonstrating tools and services which will support the ubiquity of citation links and other types of links between journals and relevant resources such as bibliographic and scientific databases.Publishers and research centres
This form of integration of journals and information resources has long been sought by the academic community. Now almost every sector of the scholarly journals publishing business is promising 'integration' achieved with links, including the largest primary publishers, secondary publishers and abstracting services and, most recently, journal subscription agents. The business of these companies is being transformed by the dramatic acceleration in the number of online journals becoming available. As one of its activities the Open Journal project has monitored this trend. From a little over 100 online journals at the end of 1995 we expect that over 3000 online journals titles will be produced in the UK alone by the end of 1997.Over the same period interest from this sector in the linking tools and services being developed by the project has risen markedly, with over 10 publishers now participating, including:
- Academic Press
- American Psychological Association (partner Stevan Harnad)
- The British Computer Society
- Company of Biologists
- Electronic Press, BioMedNet service
- Institute for Scientific Information (ISI)
- International Thomson (Chapman & Hall)
- MCB University Press
- Oxford University Press
- John Wiley & Sons
Deliverables
The publishers are contributing full-text resources and abstracts services to the development of four trial Open Journals in the areas of biology, cognitive science and computer science. Preliminary versions of each of these Open Journals are being released to specialist test users and evaluators. Due to requirements to restrict access to some commercial materials included in these Open Journals, the fourth of the journals will be compiled from openly-accessible materials in one of the above fields so that the concept can be evaluated by a wider audience. After initial evaluations these Open Journals will be re-released with more links, more features and a redesigned user interface, where necessary, that enhances link following and supports simple link authoring by users.Linking tools and services
The aim of the project is to provide powerful hypermedia linking techniques to allow users direct access from the Open Journals to secondary information resources, instead of requiring them to use these resources independently. The links are created dynamically at the users' request and do not need to be explicitly embedded in the journal papers when they are authored, thus realising the concept of the Open Journal.To do this the project is using the Distributed Link Service (DLS), which independently of the project is now available commercially as Webcosm. Instead of coding link data within the original documents, as in html, the DLS stores these data separately in link databases, or linkbases. The link data can be superimposed on a document via a proxy link server when it is viewed on the Web. To use traditional print terminology, the links are effectively added at 'run time', i.e. at the last possible moment before delivery, appearing as familiar Web 'buttons' within the text. In other words, the page will look like any other Web page, but it may or may not have many more links.
Another feature that differentiates this approach from other linking initiatives is the document formats to which the links can be applied. While html is the most popular format on the Web, pdf such as is viewed by Adobe Acrobat is currently the format most widely used by Web journals. With the DLS the project can apply links to documents in html and pdf.
The effect of these features is to transform link authoring into link publishing. In other words, responsibility for providing links and assuring the quality and integrity of links within the fast-changing Web information environment is largely assumed by the publisher, just as for the underlying text.
For the publisher, the link service provides a means to add value to journals by controlling and maintaining links, and it enables publishers to respond to the new dynamic that the Web imposes. For the user, the link service provides direct access to selected materials. In this way access to information will be faster, with less user-based navigation in and out of separate resources, and more precise, going directly to the relevant article or section of text rather than to a table of contents.
Supporting technology: software agents
Although based around the DLS, the project is also developing a framework of information retrieval technologies. Through the development of subject-expert software 'agents', users will be offered a greater range of relevant resources than they could discover alone. The emphasis is on developing agents that support resource discovery and linking. Typical applications for which prototypes have been created include a citation agent and an agent for processing the results of Web searches.The citation agent matches references within journal articles against a database of bibliographic information. As well as recognising the citation, which requires an understanding of the house style of the publication, an understanding of the document coding format and the ability to parse natural language structures, the agent has to locate the cited paper and create a link to that document. Another agent takes the results from a Web search engine such as Alta Vista and adds them to a linkbase. For example, a user could request that the results of keyword searches, after pruning, be turned automatically into linkbases.