The XML Connection

A new language for the Web will transform electronic commerce and information management.


The questions began several years ago, when Johns Hopkins Medical Center started making all of its patient records available electronically, on a new, $2 million system. Recalls Stephanie Reel, vice president for information services: "As soon as we started to populate the electronic patient records with new documents, people started saying, 'All that information is great, but why can't we have the clinic notes in the electronic patient records also?' And then, 'Why can't we have the problem reports in the electronic patient records?'"

The problem is, care providers, claims processors, and quality-assurance experts all need access to patient data, but all need different views of that data, says Reel. Right now, they don't have a simple, universal way for their computers to locate and retrieve exactly the information they have to see.

In the future, they will. It's called Extensible Markup Language, or XML. Completed in 1998 by the World Wide Web Consortium (usually referred to as the W3C), XML provides a way for content providers to add "tags" to their text, to indicate how it is being used. "What XML allows you to do is richly define the data you already have," says Joshua Walker, an analyst at Cambridge, Mass.-based Forrester Research Inc., an information technology advisory firm.

At Johns Hopkins, clinic notes, problem reports, and other information are being added to all the records in the system. The medical center expects to save $2 for every $1 spent on the system. According to Reel, all the information will be tagged with XML, which should substantially improve access to it and, consequently, improve patient care.

The situation at Johns Hopkins is a microcosm of what's happening everywhere, as more and more data is being stored and made available as content in corporate databases and, especially, on the World Wide Web. The standard markup language of the Internet, HTML (Hypertext Markup Language), uses tags to tell a Web browser how information should be presented on a computer screen. But HTML says nothing about the information itself.

XML, on the other hand, enables a computer to identify the kind of information received, whether it's a purchase order, a medical record, or a customer query. Platform independent, it seems destined to become a universal format for data exchange, which has led some observers to call it an Esperanto for the Internet. The implications for electronic commerce and communication are enormous.

Finding Needles in Haystacks

XML will transform document and information management, too. Take this article, which is also archived on CFO magazine's Web site. If you wanted to find it and other articles like it, how would you do that? A simple search of "Johns Hopkins" with an Internet search engine might turn up more than 100,000 "hits," or matches on Web pages; one of them might be this article. If you added a few extra terms to your search request, such as "magazine," you could narrow the search considerably.

But when Web documents contain standardized XML tags, the references to Johns Hopkins in a CFO story might be tagged as "a user story in a magazine article." A search engine in, say, 2005 might be able to classify "Johns Hopkins" hits according to their XML tags, or allow you to search for specific tags. Thus, retrieving only Web pages containing articles like this one would be a cinch.

Today, however, XML tags are still in the initial stages of being standardized. The process is being driven by a variety of standards bodies, industry groups, and companies. Just about every major software vendor has announced its support for the new language.

The American Institute of Certified Public Accountants is creating an XML-based specification for the preparation and exchange of financial information, which the institute expects will be accepted throughout Corporate America. This would make it possible for application software to easily recognize any financial statement and know what it means.

"Accounting is the language of business," says Barry Melancon, president and CEO of the AICPA, "and [XML tags] will make it easier to share information expressed in that language by permitting computer applications to understand our vocabulary."

Sharing information is what makes XML so powerful. "Companies need to think about what information they want to distribute and exactly which people are going to get it, with a high level of security, and [about] how that's going to change their business," states Timothy S. Sloane, analyst at Boston-based consultancy Aberdeen Group Inc. "XML lets you distribute information securely at a much lower cost."

A Replacement for EDI

Not surprisingly, electronic commerce is drawing all the attention as the most important potential application for XML, since XML enables trading partners to exchange information easily and inexpensively.

Prior to XML, the only standard way for business partners to exchange documents electronically was through electronic data interchange (EDI). The advantages and limitations of EDI are well known. It was developed in the 1980s as a way for businesses to exchange purchase orders, invoices, and other documents. In the early 1990s, EDI became a source of controversy as Wal-Mart and other large retailers forced their suppliers to use it, despite the substantial programming expense to implement each relationship, and the communications expense of transmitting the electronic documents over expensive value- added networks (VANs). But EDI has never really taken off, despite years of effort.

XML plus the Internet is a far cheaper and less cumbersome alternative to EDI and VANs. A purchase order or an invoice can be sent by ordinary E-mail as long as the various fields are identified with XML tags, thus obviating the need for complex formatting requirements.

"XML is finally delivering on what EDI promised," declares Monica Luechtefeld, president, marketing, at Office Depot Inc. Office Depot, the world's largest seller of office products, has used EDI with many of its customers for years. "Many of our customers have entrenched EDI systems, and we expect them to continue using it," says Luechtefeld. "But with new customers, we'll use XML rather than EDI."

Office Depot began the transition to XML by modifying its order-entry software to communicate with XML tags used by E-commerce software from Commerce One Inc. Any customer that uses purchasing software from Commerce One can place orders with Office Depot and track them electronically.

"We started with Commerce One because they were the first to implement XML," says Luechtefeld. "But Ariba and IBM will be ready very soon, and we'll be partnering with them, too."

While Office Depot is using XML to facilitate E-commerce with its customers, W.W. Grainger Inc. is using XML to communicate with its suppliers. Here, the situation is a little different. W.W. Grainger, the nation's leading provider of MRO (maintenance, repair, and operating) supplies, needs to have detailed product descriptions on its Web site, and it has to obtain those descriptions from its suppliers.

Grainger is using CenterStage, from OnDisplay Inc., as middleware between the Grainger database and the databases of its suppliers. The software, which the vendor says typically costs between $75,000 and $150,000, based on services and other factors, automatically collects product descriptions from vendors' databases and legacy files; adds XML tags to those descriptions, following rules in schemas that Grainger supplies; and stores them in Grainger's database. The tagged data can then be displayed to Grainger's customers when they request product information on the Web.

The bottom line? "This has cut in half the time it takes to post new product information to the Web catalog," reports Ron Paulson, executive director of product information and sourcing at W.W. Grainger.

Developing a Vocabulary

Is there anything that you should be doing about XML today? Certainly, you should start arranging for all documents produced by your company to contain XML tags.

Esperanto or not, however, make sure up front that your tags will mean the same things as the tags your trading partners will use.

"Identify the partners that you need to exchange information with," advises Joshua Walker of Forrester Research, "and start using XML in your exchanges. Start developing an XML vocabulary, the standard set of tags that you'll be using in your company, and talk to other companies in your industry to develop a common vocabulary. You have to agree on what those tags are going to mean--that's what makes XML so valuable."

----------------------------------------------- --------------------------------- How XML Works
Suppose a used-car Web page displays product information like the following: Blue Honda Civic $21,995

This ordinary text is baffling to a computer. Is "Blue" a color, or the name of a manufacturer? Most computer applications wouldn't know without XML tags. So here's what it takes to make this line of text machine- readable.

The first XML tag (bracketed in angles) simply establishes that this is a product:
Blue Honda Civic $21,995

But what kind of product? Blue Honda Civic $21,995
This more-complex XML tag says that the product is a car.

Finally, let's embellish to the max, by describing every word with an XML tag: Blue Honda Civic $21,995

In five years or so, you'll be able to use an Internet search engine to make a query such as, "Give me any Web pages that reference blue cars from Honda," and the search engine will be able to find them, thanks to the XML tags.

(Note: When your Web browser displays this text, it ignores the tags. Your computer won't show them unless you command it to do so.)

----------------------------------------------- --------------------------------- Who's Behind XML?
Here are a few of the many entities that are developing XML standards or integrating XML with their products and services:

Standards Bodies Consortium of companies working on XML E-commerce standards Nonprofit organization dedicated to advancing industry standards Consortium to create XML- based trading processes for IT supply chain Internet standards body responsible for approving XML innovation

Software Vendors

IBM: Leading in XML parsers and XML style sheet (XSL) innovation
Microsoft: First to deliver XML browser support and "save as XML" in Office 2000
Oracle: Strong XML support in database and iFS
Sun/Netscape: XML extensions to Java and XML support in NetDynamics this year

Industry Giants

AT&T: Working with Lucent and Motorola to promote voice XML for telephones
Charles Schwab: Leading the financial industry's move from OFX to XML-based IFX
The New York Times: Demonstrating XML's syndication power with AdTimes/2000

Source: Forrester Research Inc.

----------------------------------------------- --------------------------------- XML+LDAP+PKI=Universal Access Control
Extensible Markup Language (XML) is a fundamental technology, with virtually unlimited applications. According to Boston- based Aberdeen Group Inc., one of the most powerful applications for companies engaging in business-to-business communications with customers and vendors is to combine XML with two other technologies.

One is lightweight directory access protocol (LDAP). LDAP is a company's library catalog, providing a directory of all content in all databases and documents.

The other technology is public key infrastructure (PKI). PKI uses encryption and digital certificates so that when data is sent, the sender is guaranteed that only the named recipient can read it, the recipient is guaranteed that no one but the named sender could have sent it, and everyone is guaranteed that the message wasn't "hacked."

Each technology becomes even more powerful when combined with XML. If the company's documents contain XML tags, then the LDAP library catalog becomes deeper and more complex, as it cross-indexes the tags with the documents. Furthermore, if each document contains an XML tag that indicates whether the document is confidential and who should be allowed to see it, then access information can go into the LDAP catalog as well.

Now throw in PKI encryption. Using PKI, the LDAP catalog can verify the identity of the person requesting the data, and that he should be permitted to see the documents requested. Putting all three technologies together allows a very flexible environment in which all content in the company is automatically cataloged and cross-indexed, and is available over the Internet to anyone in the world, but with full access controls and authentication.

According to Aberdeen analyst Timothy S. Sloane, the adoption of these technologies will transform business-to-business relationships during the next five years, just as the adoption of HTML transformed business- to-consumer relationships during the last five. "The cost of the creation, maintenance, and enforcement of trust relationships will drop significantly as companies deploy directories that can be shared across the Internet," predicts Sloane. "Once information is consolidated in the directory, it can be more easily controlled and accessed, enabling the enterprise to react more quickly to the needs of the market, while also reducing the effort to implement new business processes."


Read next:

Finance At BeyondCore