CITATION LINKING TO EJOURNALS List: Hyperjournal 14 Oct. 1995 From Archie Zariski, School of Law, Murdoch University, Western Australia [cut] Our position is that (a) citation (to an ejournal paper) accurately reflects the fact that the journal is electronic and by analogy with other journals the citation is not intended to point to the location where the journal may be found. (This in our view is primarily the responsibility of libraries and legal research assistance.) 15 Oct. 1995 From Tony Barry, Head, Center for Networked Access to Scholarly Information, Australian National University Library One could take the contrary view that the link is the defining item in this environment (or will be if URN's are ever developed) and therefore the citation is not needed as the link takes you direct to the material quoted. Rather than needing a footnote to a citation to find the document you link direct to the material [cut} The citation is needed to help people find the document which is your authority for a statement. It is a description which you can take to an external index or a catalogue which has sufficient detail so it can be located and identified as correct when eventually found. If the document is a mouse click away why is the citation needed other than for reasons of tradition? Tony __________________________________________________________________________ Tony Barry Centre for Networked Information and Publishing & also Centre for Networked Access to Scholarly Information Australian National University Library Canberra A.C.T. 0200, AUSTRALIA tony@info.anu.edu.au17 Oct. 1995 17 Oct. 95 From Steve Hitchcock, Open Journal Project, University of Southampton Archie Zariski wrote: >Our position is that this >citation accurately reflects the fact that the journal is electronic and >by analogy with other journals the citation is not intended to point to >the location where the journal may be found. (This in our view is >primarily the responsibility of libraries and legal research assistance.) Isn't this lack of a pointer rather a waste in an electronic journal? It's useful and easy to implement. I'm with Tony Barry that the link is one of the defining items in this environment. In the Open Journal project such links don't have to be added by the author, they can be added later by a publisher or editor, or by a library if you wish, or all of these (you don't need to change the html source code to add links). Admittedly URLs can disappear, but if you are aware of a change the link is easily modified. I'm amazed that the majority of refereed e-journals, both commercially produced and non-commercial, even when they are in html, have no button links at all, not even citation links. It can't be said that there is too little to point at. Inter-journal links are more difficult in pdf, but the CAJUN researchers at Nottingham University have developed a partial solution even here and we expect to be able to apply 'open' links to pdf documents more effectively in due course. Steve Hitchcock Open Journal project Department of Electronics and Computer Science University of Southampton SO17 1BJ, UK sh94r@ecs.soton.ac.uk From Damien Keown, Hyperjournal list owner Steve Hitchcock wrote: >In the Open Journal project such links don't have to be added by the author, >they can be added later by a publisher or editor, or by a library if you >wish, or all of these (you don't need to change the html source code to add >links). This is a great idea. Could you give a few details of how it works, exactly? Is the software available now or still under development? Thanks, Damien Keown d.keown@uk.ac.gold (Damien Keown) 19 Oct. 1995 From Steve Hitchcock Yes, the software is ... still under development! The software is called the Distributed Link Service [cut]. We recently took up space on some lists announcing the project, so I won't bore people with the technical details here. Suffice to say that while the underlying client-server implementation and mechanisms may seem rather complex to some, the point is that once it's set up it [will be] easy to use to make links via mouse and window operations on multiple platforms - PC, Mac or Unix. [Those who are interested can find the paper 'The DLS: a tool for Publishers, Authors and Readers' at ??????. The first section sets the context and may be most helpful, before it becomes more technical (the paper was written for Web developers)]. The essence of the Open Journal project is not to develop the DLS software - that is a parallel development - but to apply it in the specific context of electronic journals. This is already being done but the aim is to improve both the speed at which links can be created on a publishing scale, through automation, and the quality of links made in this way. The project is also developing ... links [for] multiple formats, notably in Acrobat which will be the principal format used by the huge number of commercial scientific e-journals that will appear next year, as well as in html. We are not talking here just about links on references, or intra-article links to tables or figures, which is the best most e-journals can offer at the moment. In principle any word or feature within a text can be the start of a link to a related document. One of the journals we are working with is in the field of biology where the richness of online resources, databases and other biology journals, almost makes this possible. I have saved the best to last. The links made by either publishers, authors or readers are stored in link databases, or linkbases. Using a linkbase the user can follow a link from the original document to another document, and regardless of the origin of the second document can follow a link out of that document to a further document, and so on. If this doesn't sound so great, try following at random the buttons in standard Web documents and see how many documents you can visit before reaching one that has no further links to follow. Many html documents have no links, and pdf or ascii documents invariably don't have links. Another test is to find a Web document that interests you and identify any words or phrases that, if it were possible, you would want to know more about. On how many of these features has the author anticipated your interest by coding a button? The linkbase can be created to put links into such documents, where links don't exist or where additional links are required. With the development of Web metadata services, the wide availability of search mechanisms and a link-rich environment, not to mention the advent of personalised intelligent agents, users need never get lost in the Web again. In practice, we shall see. Steve Hitchcock Open Journal Project Multimedia Research Group, Department of Electronics and Computer Science University of Southampton SO17 1BJ, UK sh94r@ecs.soton.ac.uk 19 Oct 1995 From Dr D.McL. Roberts, Dept. Zoology, The Natural History Museum, London Steve Hitchcock wrote: >In the Open Journal project such links don't have to be added by the author, >they can be added later by a publisher or editor, or by a library if you >wish, or all of these (you don't need to change the html source code to add >links). Admittedly URLs can disappear, but if you are aware of a change the >link is easily modified. It may be easy, but it is also going to be very frequent over all e-journals and Web pages. So who will actually do this easy task? We are still handling e-publication like paper, but without the commitment of library back-up. Of course e-journals should sport links and other hypertext features, but I'm with Archie in thinking that this is the role and responsibility of libraries not publishers. Cheers, Dave -- Dr D.McL. Roberts, Dept. Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD Great Britain dmr@nhm.ac.uk 20 Oct. 1995 From Steve Hitchcock Dr D.McL. Roberts wrote: >... I'm with Archie in thinking that this is the role >and responsibility of libraries not publishers. No, it is not just libraries that should be responsible for maintaining links. In an earlier reply I described a new way of applying links that is being developed by the Open Journal project, in which I said: >The links made by either publishers, authors or readers are stored in link databases, or linkbases. I should perhaps have emphasised that these linkbases are quite separate from the Web documents they might point to. Neither are linkbases Web documents themselves: the result of a WebCrawler search, for example, is a list of button links, but this is not a linkbase. Linkbases are a hypertext application, which by definition (according to Conklin, IEEE Computer, September 1987) requires an underlying database, sophisticated machine-supported links (e.g. single-click button links, but not filename typing) and an interface for online navigation between documents. In this context, link data from a link database relating to a particular Web document is overlaid on that document when it is viewed, identifying links on the document as, but not necessarily as, blue buttons. If users create their own linkbases, the linkbases may be relatively small, and they are a personal resource that the user can maintain. When overlaid on a Web document a personal linkbase may give that user a unique view of that document. Some very large linkbases may have wider application and might be 'published', i.e. made more generally available. It would make sense in this case for the linkbase to be updated by the publisher, in the same way that publishers issue new editions of textbooks, for example (although linkbases would hopefully be updated rather faster). Librarians could develop and maintain linkbases that are relevant to a library's resources, and this is another application. The textbook is a useful illustration. Rarely does a textbook contain entirely new knowledge, but it is usually a new view, the author's view, of the underlying body of knowledge. Linkbases are another way of imposing a view on a base of knowledge. This may be an author's view, an editor's view or, maybe more commonly in the textbook case, a lecturer's view on an array of resources (available electronically of course). Who should take responsibility for maintaining the textbook linkbase? The developer. If linkbases prove to be useful, vast numbers of them will be created of different sizes for different purposes by different people for different uses. The question of who maintains a linkbase will relate to these issues and is therefore not the exclusive domain of librarians. Steve Hitchcock Open Journal Project Multimedia Research Group, Department of Electronics and Computer Science University of Southampton SO17 1BJ, UK sh94r@ecs.soton.ac.uk