I'm developing a website of text, digital images, and references. Should I be making special efforts to implement any kind of reference linking?

First, be clear about what reference linking is. It is the process by which a bibliographic reference or citation to a document becomes a link to that document. How does it differ from a link on a Web page to another Web page? The answer is that, at its simplest, it does not. An HTML link is an in-line reference link. Sometimes a set of links is gathered at the end of a Web document, reproducing the format of traditional, non-digital journal articles, but the placement of the links in the document is a convention, not a necessity. The same is true for links in PDF documents or in other digital formats. Another convention from the print world is the format that displays metadata about the referenced document—author, title, publisher, and so on. In non-digital documents, these elements of a bibliographic citation help a reader find the cited work. In a reference-linked digital document, the link "finds" the document for the reader. Thus, reference linking is a value-added feature of digitized materials.

Second, decide whether the users of this material will find reference links useful. Assuming the answer is "yes," you are still faced with several important questions. Does the community of users have permission to view the cited documents, and does the mechanism used to link to the documents recognize those rights? Some of the documents you might want to link to could have access restrictions that exclude some potential users from viewing the documents. Depending on the domain of the collection and the cited documents, you could be providing users with many links that end in refusal. The "appropriate copy" problem, also known as the "localization problem," occurs when a document is available from multiple sources, some of which may not be available to a particular user. In this case, the link you create will provide some users with access but will exclude others who rightfully should be able to view the document from a different resource provider. To solve this problem, an institution can provide localized links using a tool like SFX, one product that has been fairly widely used for this purpose. (1)

Third, you need to consider the stability of the links. Are the cited documents located at a stable address? A permanent address is an unrealistic wish in the fluid digital world. The only way to make sure that an address points to a document is to change the address whenever a document moves. If this change had to be made for every link that pointed to a particular document, the cost would be such that few people would ever bother to use reference linking. Some sort of persistent identifier is needed to decouple a document from its current location. Several systems of redirection are now in use—PURL servers for local collections, Digital Object Identifier resolution services, and persistent identification numbers assigned by reviewing bodies such as Mathematical Reviews or e-print services such as the e-print ArXiv, among others.

Fourth, you need to consider if there are sufficient resources available to create and maintain the links. Can you afford to add this value to the collection, and can you ensure the maintenance of the links' integrity? Adding reference links has a cost, whether the links are created manually or automatically. There are trade-offs in using either method. Humans are relatively slow and expensive, but they parse references into semantic constructs that can be sent to a lookup service easily. Automated methods are usually fast and inexpensive—for large numbers of links—but are not reliable in making decisions when ambiguity exists. Moreover, automatic linking processes are inexpensive only during operation, but not during the programming phase. General linkers that take a document and find the locations of every referenced document do not yet exist.

Once the elements of a query to a lookup service are in hand, there can still be costs involved in finding a location to link to. Using a standard Web search engine takes time. The current economic model of sending metadata to a DOI resolution service involves payments during the registration and lookup phases (but not when a user clicks on a link). Behind every persistent identifier is some infrastructure for updating links. The costs are sometimes carried by the document owners, sometimes by the lookup services, but in every case there is a cost that you should consider before you make a decision about reference linking a collection.

In summary, reference linking is a potentially important value-added feature to digital documents. The decision to add that value depends on your users, the nature of your collection, and your budget.

Errata: Clarification note added 17 October 2002: Original text named SFX as "the mechanism" to provide local links. SFX is an example of an implementation of the OpenURL technique for reference linking; other tools also exist.

Publishing Information

RLG DigiNews (ISSN 1093-5371) is a newsletter conceived by the members of the Research Libraries Group's PRESERV community. Funded in part by the Council on Library and Information Resources (CLIR) 1998-2000, it is available internationally via the RLG PRESERV Web site. It will be published six times in 2002. Materials contained in RLG DigiNews are subject to copyright and other proprietary rights. Permission is hereby given for the material in RLG DigiNews to be used for research purposes or private study. RLG asks that you observe the following conditions: Please cite the individual author and RLG DigiNews (please cite URL of the article) when using the material; please contact Jennifer Hartzell, RLG Corporate Communications, when citing RLG DigiNews.

Any use other than for research or private study of these materials requires prior written authorization from RLG, Inc. and/or the author of the article.

RLG DigiNews is produced for the Research Libraries Group, Inc. (RLG) by the staff of the Department of Preservation and Conservation, Cornell University Library. Co-Editors, Anne R. Kenney and Nancy Y. McGovern; Production Editors, Martha Crowe and Barbara Berger Eden; Associate Editor, Robin Dale (RLG); Technical Researchers, Richard Entlich and Peter Botticelli; Technical Coordinator, Carla DeMello; Technical Assistant, Kimberly Gazzo.

All links in this issue were confirmed accurate as of October 11, 2002.

Please send your comments and questions to RLG DigiNews Editorial Staff.