XML for Content and E-Commerce.

A striking advantage of computer-based data is its flexible re-use for multiple purposes. Re-purposing digital document content creates new publication avenues and potentially creates new revenue streams for competitive enterprises. The ability to transmit data faster and more economically by electronic means creates incentive for using computer-based methods of document distribution instead of more traditional "print and distribute" models. However, most organizations investing time and resources in creating document content want to get the greatest possible return. If a single investment in content creation yields multiple products and services, operating cost is reduced and new income generated.

The use of extensible markup language (XML) meets these goals admirably. Document data and text content encoded with XML can be posted to Internet Web sites, processed with desktop publishing software for distribution on CD-ROM media, and used in e-commerce applications. In fact, the growing acceptance of XML use for building document content is creating a sense that XML is a new open data standard that can be used for data exchange, document publishing, and eventually document archiving. Organizations have begun to employ XML for many document publication and data exchange purposes.

As with most new technologies, implementation realities result in scaling back expectations as "lessons learned" begin to surface. Some organizations will experience XML shock as they retool information systems departments and computer infrastructures. Others will find that XML documents are actually more limited than those produced in more traditional formats and that not all Web-site-hosted documents really need XML to be informative and functional. XML use for content management has enormous potential, but it must be implemented realistically.

Evolving Content Management

The explosion in Internet use during the 1990s and the subsequent drive to Web-enable many computer software systems resulted in the need for a standard method of presenting information on Web sites. For many years, basic-level hypertext markup language (HTML) was used to present textual content across the Internet by creating tagged text documents. Tags -- special text codes -- were placed around various text and graphics elements in the document to indicate structure and specific formatting when the document displayed or printed. HTML was a convenient subset of standard generalized markup language (SGML), an international standard that had been around for many years.

SGML is maintained by the International Organization for Standardization and is not bound to any particular computer operating system or hardware platform. Originally intended to foster the exchange of text-based data between organizations, SGML is a language syntax used to create markup language tags for different types of documents so that their content and formatting can be dealt with separately. Having a standard language for managing document content has proven advantageous for organizations over the years as a basis for content communication and exchange. For these reasons, there are SGML-based text editors, typesetters, and document management products, and many organizations have placed document content in a standard format by using SGML.

Since SGML documents are not tied to any one vendor or software, various industries have created their own document types (document type definitions -- DTDs) for information commonly exchanged in their business activities. However, SGML requires special training to use, and it is not particularly fast or efficient in the way it runs on computer systems. In addition, all documents must use a DTD designed by the document's creator.

As a subset of SGML, HTML is useful for publishing documents on Internet Web sites and does not require DTDs to be created. Internet browsers, such as Microsoft's Internet Explorer and Netscape's Navigator, display HTML well and are relatively inexpensive. HTML is easy to learn, and the coded hypertext links among HTML pages form the basis for much of the Internet's fast and seamless document accessibility. HTML's simple, inexpensive nature contributed to the vast proliferation of HTML-based document content on the Internet almost overnight.

As HTML documents have become common, their inadequacies have become more obvious. Pages often print poorly due to the lack of sophisticated page-formatting capabilities. HTML does not work well with text justification, kerning, and hyphenation, or with features such as hanging indents and multiple columns that word processors easily address. HTML has no method of creating custom tags for specific document types, and HTML tags do not have good structural relationships, thus making it difficult to present content in rigorously structured pages. For these reasons, many contemporary Web sites use HTML for visually interesting and informational page displays, but make actual documents downloadable in Adobe's portable document format (PDF). Electronic documents in PDF appear the same way across all computer platforms, software systems, and printers. In addition, the Adobe Acrobat Reader software is freely distributed, basic PDF files are inexpensive to produce, hot links to other documents or Web sites can be embedded, and there is little training required.

PDF also has a few drawbacks. It is possible to create page-oriented hot links internally within PDF files. However, they have minimal structural hierarchy and content indexing ability. Text searching is limited largely to the facilities provided by Adobe software products. Most information retrieval is page oriented, as are most document navigation and display tools, due to the paper-replica orientation of the Adobe environment. A PDF document's precise visual display can vary slightly from the original document, depending on the available character fonts and color palettes of computers or printers. Software tools are unable to precisely access textual or numeric data within PDF files, so PDF files are not especially useful for document data exchange.

XML, in contrast, excels at content management. It is not limited to a fixed number of tags, as is HTML. XML does not require using a DTD, plus new elements and attributes for use within documents can be defined. In addition, XML can create custom tags such as for improving the recall specificity of text searching. It also can be used to define new markup languages to customize data management when building complex documents, thus ensuring that the data can be used by more than one software application.

Instead of being presentation oriented, as with HTML (in browsers) and PDF (in print and viewed by Acrobat Reader), XML enhances document content management while still enabling a variety of software programs to handle presentation and publication. Information content can be stored in an XML document and then processed with HTML, a word processor, Adobe Acrobat, or database software.

XML Rockets into Cyberspace

As with all new technologies or methodologies, there is a great deal of hype with respect to XML. The advantages of XML document content management seem so overwhelming that many organizations are moving rapidly to embrace it.

There are few business environments where the exchange of information internally and externally is more important than in government settings. A U.S. federal government chief information officer council has now created an XML working group at www.xml.gov. The group promotes XML use by government agencies, encourages the learning of XML, and fosters sharing of implementation experiences. The U.S. Defense Department is creating a similar site at www.xml.mil, while the General Services Administration, the U.S. Fish and Wildlife Service, and the U.S. Patent and Trademark Office (PTO) already have internal XML initiatives (O'Hara 2000).

The Federal Deposit Insurance Corp. (FDIC) plans to separate content from format when it publishes its quarterly Regional Outlook to the Web using XML. Being able to multi-publish this description of economic conditions in both paper and Web formats is a big incentive. However, the FDIC also exchanges data with the Federal Reserve Board, the Office of the Comptroller of the Currency, and private sector banking companies, where XML is expected to be especially useful (Caterinicchia 2001).

The National Archives and Records Administration (NARA), concerned with the long-term preservation of documentary information, is also entering the XML fray. An electronic records archives program will use XML to ensure that documents can be read in the future without regard to the software used to produce them. Some research is under way to ensure that records' authenticity is preserved, and this research is expected to show that more informational constructs will be required than just DTDs and document style sheets. In addition, XML topic maps for knowledge representation and communication may prove beneficial in helping NARA connect records to various agencies' business processes (O'Hara 2001).

The PTO has set up an electronic filing system that allows it to accept patents across the Internet in XML format using public-key-encryption-based digital signature technology (www.uspto.gov/ebc/index.html). The PTO offers a downloadable patent application specification authoring tool (PASAT) that works with word processors to export documents in XML format. SGML is now used for publishing patents, but the agency will switch to XML in January 2002. One incentive for XML-based electronic publishing is that the agency will move to a new building in 2003 where there is little space for paper files (Daukantas 2000).

In the private sector, e-commerce is driving XML use, especially for data interchange. As Web sites began to take orders for commodities during the 1990s, it soon became clear that some document-based information would be dynamic by its very nature. Lists of available products, current inventory data, regional sales data, or an organization's roster of current members may fluctuate by the minute. To deliver this type of information across the Internet has previously required the integration of database software applications with a Web-based interface. Now, with XML, organizations can have Web-based data processing operating in the background to ensure that updateable documents are accessible. Although HTML simply indicates how data should look, XML can tell an application what the data means. For this reason, the same data can be displayed differently to different users and in disparate computer applications. In some cases, XML implementation replaces electronic date interchange (EDI) applications.

EDI's advantage as a business method has always been the ability to conduct very fast, reliable, and secure transactions over value-added networks (VANs). For organizations conducting thousands of business transactions a day, EDI was the best solution for automating the business process. VANs were not inexpensive, however, as they were primarily dedicated, leased computer communication lines. In addition, EDI software and expertise are required, and most EDI transactions are one-to-one business communications. In contrast, XML data formatting knowledge and expertise are freely available, XML is customizable, and it can be used in a one-to-many business model -- the model of the Internet. For these reasons, many organizations plan over time to replace some EDI transactions with XML-based business processes (DeJesus 2001).

After all, how can they go wrong when XML use is a recommendation of the World Wide Web Consortium (www.w3c.org) with 400 international members dedicated to Web interoperability, and an XML working group that makes speeches, presentations, and technical specifications freely available at www.w3.org/XML? In addition, the Organization for the Advancement of Structured Information Standards (OASIS), a non-profit, international consortium that creates interoperable industry standards based on public standards like SGML and XML, has a number of XML-related committees at work (oasis-open.org/committees/committees. shtml).

Additionally, many organizations now offer consulting and education in XML. For instance, the O'Reilly publishing company produces the Web site XML.COM (www.xml.com) with resources, buyer's guides, answers to frequently asked questions (FAQs), and free newsletters. Publications such as Web Techniques contain regular columns covering XML (e.g., XML@Large), and most major computer industry trade journals feature articles covering XML (see references).

Organizations such as the Graphic Communication Association are educating their members regarding XML's impact (www.gca.org/ whats_xml/). The variety of conferences, seminars, and publications are almost overwhelming. A visit to any major bookstore will reveal 10 to 20 titles of XML-related textbooks, technical references, and self-help manuals sitting on the shelves, ranging from The XML Bible with CD to XML for Dummies. A quick search for XML-related topics on the Internet book-seller Amazon.com's Web site indicated 354 related items for sale, of which two-thirds of the first 50 items have XML in the title.

Content = Records?

The advantages of XML will undoubtedly make it a leading development platform for both document creation and data exchange; but there are factors to consider from both the records management and business perspectives to temper assumptions that XML will invade all document-based applications. As with all new technologies, some education, infrastructure retooling, and capital investments will be required, leading to questions about return-on-investment and the value added to current business processes.

An unforeseen benefit of interest in XML is a renewed interest in data standardization and integration across enterprise computing architectures. However, as in the past, organizational politics and the technical intricacies of legacy systems can frustrate any attempt at achieving cross-organizational data sharing. Most important is the renewed realization that "an enterprise understand and resolve the different words and meanings it uses to refer to things important to it" (Finkelstein 2000). Records and information management professionals must seize this opportunity to point out that data standardization, commonly accepted names for records series, and integrated records classification and filing systems are all strongly related. How can a XML or SGML document type definition not be related to a records series?

If the goal of XML is re-purposing so that one set of content can be published in multiple document forms, how will that affect the authenticity of documents that are declared to be records? "Form (layout and design) must be separated from content (primarily text). Although the words and pictures of a print layout are related, they don't necessarily work together the same way on the Web or on a CD-ROM as they do in print" (Ward 2000). How does one designate an official record among a set of transitory (but related) documents in multiple formats derived from the same set of data? How does one preserve the context, as well as the content and presentation of the record?

Text and graphics based on the same content can appear very different visually depending on the final publication media. Printed documents can take advantage of many color palettes, image resolutions, and media to enhance text and graphics presentation. However, when the same text and graphics are published using XML on a Web site or on CD-ROM media, a more limited array of presentation tools is available due to software and hardware limitations.

Which media format will provide the best and most authentic official record for long-term preservation of informational content? Will a record's evidentiary value be affected even though its information content is less questionable? How critical is presentation and context to the authenticity of records? It may be up to the courts to decide.

XML has much potential to save us from the continuing Towers of Babel that we create with new renditions of technology. Whether or not it will do so depends on our making well-considered decisions about how we use this new technology.

REFERENCES

Caterinicchia, Dan. "FDIC Simplifies Publishing with XML." Federal Computer Week. 8 January 2001:20.

Daukantas, Patricia. "PTO Starts E-government Shift." Government Computer News. 20 November 2000:1,41.

DeJesus, Edmund X. "EDI? XML? Or Both?" Computerworld. 8 January 2001:54-56.

Finkelstein, Clive. "XML Is Not a Silver Bullet." DM Review. October 2000:42.

Harold, Elliotte Rusty. The XML Bible. Foster City, CA: IDG Books Worldwide. 1999.

O'Hara, Colleen. "Future Electronic Records Archives Bet on XML." Federal Computer Week. 8 January 2001:24.

--. "XML Portal in the Works." Federal Computer Week. 4 December 2000:10.

Tittel, Ed and Frank Boumphrey. XML for Dummies. Foster City, CA: IDG Books Worldwide. 2000.

Ward, Noel. "The Challenges of Re-purposing." Electronic Publishing. December 2000:38.

John Phillips, CRM, is the owner of Information Technology Decisions, a management consulting firm. He has more than 20 years' experience in information resources management, specializing in automated records management systems and other technology-related areas. He can be reached at [e-mail].