Digital Object Identifiers: Not Just for Publishers
By David Sidman at 2002-03-31 00:00:00 |
The Web has often been compared to a vast jungle of information, where it's difficult to find what you really need, and when you do, it's often no longer there the next time you need it because the URL may be obsolete. Consider the two figures below:
If the Internet is to become a truly "grown up" medium for B2B transactions, commerce, and discovery and access of high-value information of all kinds (including information about physical products), it needs a way of finding and accessing information which is more permanent, reliable, and machine-interpretable than a Web which is founded entirely on connecting to things based on where they live, as opposed to what they are.
Such was the reasoning of one of the primary architects of the Internet itself when he designed the "Handle System" that underlies the Digital Object Identifier ("DOI"). Dr. Robert Kahn, co-inventor of TCP and builder of the ARPAnet, developed a system whereby every "information object" receives a unique and universal ID number (the DOI). A global distributed directory (similar to the DNS system that routes domain names) routes people or programs permanently to the object itself, as well as to any related services, transactions, or related information that the object's owner wants people or programs to be able to get to in a permanent, reliable manner.
Unlike a URL, which "breaks" the moment a website's sitemap is reorganized, or the item is moved to a different server, the DOI ensures permanent links, because if the URL for something changes, the owner of the object simply updates one central DOI record. This change populates a global directory, and from then on the directory re-points all links referencing that DOI to the correct place. This way, the owner doesn't have to find all existing URL-based links which might exist in the world, and change all of them because each one is based on a hard-coded -- but outdated -- address (URL). The chart below illustrates why the DOI assures permanent links.
The "objecgt" in Digital Object Identifier can be literally anything the owner wants to define:
- a physical product, where the DOI might point to its product information record in the manufacturer's catalog; note that it can be an individual product (i.e., not just a particular brand or model, but literally each individual product at the serial number level -- where the DOI would then point to its sales record, its warranty status, its servicing history, etc.);
- a book or magazine article or research report; it can even be a piece of music;
- a medical record, where -- besides offering security/confidentiality benefits which are outside the scope of this article -- the DOI would help satisfy the federal HIPAA requirements by pointing to all related records, whether they're at the hospital, the insurance company, etc. -- and whether or not they get moved around the Internet over time.
The DOI can also be applied at any level of "granularity" -- in the case of a product manufacturer, it could be applied not only to a complete product, but also to every part or component; in the case of a publisher, it can be applied to a whole book but also to every chapter, every section, even every illustration or photo or table; in the case of music, every individual track as well as a CD collection.
Among its many commercial benefits, the DOI can greatly facilitate (and reduce the cost of) the marketing and sale on the Web of more "granular" or "recombinant" products -- for example, selling someone just the portion of a travel guide covering the city they're travelling to, or being able to mix-and-match individual components into a customized package servicing the customer's direct, individualized needs.
Some live examples of these applications can be viewed on the Web at http://www.contentdirections.com, under "See the DOI in Action."[1]
For more technical information about DOIs, visit http://doi.org.
Beyond Publishing
Applications of the DOI are as diverse as the economy itself. It was first applied within the publishing industry, where 101 international Scientific Journal publishers have assigned more than 4 million DOIs to all their Journal articles and are using it to cross-link every footnote with every cited article, even if it's on another publisher's server.
But applications go well beyond the Scientific Journals industry, and indeed beyond publishing. Content Directions, Inc., the first commercial company authorized to offer the DOI Registration services whereby companies register their DOIs into the global DOI directory, is rolling out the DOI to the rest of the publishing industry (where McGraw-Hill recently declared its intention to be the first major publisher to implement the DOI across its entire publishing program), and has already gone beyond publishing to work with companies in industries as diverse as stock images/photos, investment banking analyst reports, museums, telecommunications equipment manufacturers, and catalog producers.
One of the strongest appeals of the DOI for industries beyond publishing is a feature known as the "MultiLink," whereby an object's DOI can connect people, websites, and even automated systems not only to the object itself, but to associated services or related information. Even if a URL could somehow be made persistent, and even if this could be done in a manner that would scale to the point where there are literally quadrillions of objects being linked via the Internet without any performance bottlenecks (as Dr. Kahn's system permits), the URL can still only point to one place at a time. By contrast, a DOI can point to all the resources the owner wants it to point to, via a single pop-up menu delivered to the user via the global DOI directory whenever the user clicks on a DOI-based hyperlink.
These multi-link options can point to any and all services that might benefit the owner's customers, its business partners, and even its own internal object management needs. For example, a manufacturer can use a product's DOI externally to connect its prospective customers to all the places they can buy the item, or all the locales they can get it repaired, or to the manufacturer's own website to learn more about the product, or to places where the customer can buy related products, components, or accessories.
That same manufacturer can also use the DOI internally to link together all the internal systems which might contain information about that product: e.g. the product specs housed in one system, the marketing information housed in another system, the warranty information housed in a different system. Some of this information can also be exposed to the company's business partners (in a secure/confidential way): e.g. the DOI could multi-link a retail partner to the shipping status for that product at the manufacturer's warehouse, or to pricing information, or to the order entry system.
Here again, all these links never break, because if the manufacturer needs or wants to change any of these target destinations that the DOI points to, the manufacturer just updates one central DOI record rather than finding and fixing all the individual links throughout the web and other media.
What it means for CMS and DAM vendors
As the DOI spreads from publishing to other industries and becomes (over time) the primary mechanism by which people find and access structured information on the Internet, I believe that every company in every industry will want to assign DOIs to their objects. This represents a tremendous opportunity for CMS and DAM vendors.
By incorporating DOI support into CMS/DAM systems and thereby facilitating DOI implementation on behalf of their own customers, CMS/DAM vendors can greatly improve the attractiveness of their products. "DOI support" can be as simple as adding a new identifier field (a "DOI" field) to existing content database schema (or modifying the existing content ID field that already exists there, so that it can now accept a DOI in addition to whatever other ID numbers are already permitted). This way, when the customer adopts the DOI there is already a suitable field to be populated.
Or the vendor can go further and work with a DOI Registration Agency ("RA") to integrate their CMS or DAM system directly into the registration facility run by the RA. Even this is not terrifically complicated: it is primarily just a process of exporting the product metadata from the CMS/DAM system to the RA on a regular basis, so that the customer can easily register and maintain the DOIs within the global DOI directory via the RA's registration/maintenance system.
Another benefit for CMS/DAM vendors is that once their customers implement the DOI, they are likely to want more advanced features in the CMS/DAM system that will permit them to manage their products on a more granular basis, to mix-and-match different products together, and to otherwise drive a whole new level of sophistication within their CMS/DAM system. The demand for these features will allow CMS/DAM vendors to greatly increase the feature-sets that they offer, and perhaps correspondingly derive additional revenue from their existing customer base.
Alternative to Centralized Enterprise Repositories?
On an internal asset management level alone, the DOI provides a way that CMS/DAM vendors can deliver much higher value without increasing their customer's costs to deal-breaking levels, and without straining the capabilities of today's CMS/DAM products. For much of the last 5 years, some vendors have pursued a pipe-dream of enterprise-wide content management via a single huge, centralized repository housing all information assets. But instead of literally centralizing all the assets, it is now possible to centralize only a "virtual directory," where a lookup database allows an internal user to find any asset they want via a search based on the asset's metadata. Then when the user finds the listing of the asset, its DOI can link them in a single click to any related information they need.
In a publishing scenario, the DOI could link to the native content files for retrieval and re-purposing, the licensing system to determine whether the asset is fully owned by the publisher and if not what restrictions might apply, or to the marketing database where promotional material is kept.
For an equipment manufacturer, the DOI could like to the specs, to the manual, to the warranty options or servicing plans available, or (if implemented at the level of each individual instance of the product) to the actual sales records, current warranty status, servicing history, and so forth. In all these cases, it would have been cost-prohibitive and highly impractical to literally centralize all this information, whereas it is fairly cheap and easy simply to maintain a central DOI record which points to all the relevant information in all the different systems where that content may reside.
Then beyond the company's internal asset management requirements, the DOI can help streamline operations and reduce costs throughout the external supply chain or distribution/sales chain, by virtue of providing a unique and universal ID number which allow all parties' computer systems to talk to each other, just as the UPC or bar code does this today in the physical world.
The Universal Product Code (UPC) is used as a unique product identifier
by all parties throughout the supply chain. (Click to enlarge image)
When CMS/DAM vendors begin to enable all these benefits by facilitating their customers' creation and maintenance of DOIs, they will further integrate their products into the core operations of their customers.
The DOI unifies the supply chain or distribution/sales chain in the digital
world, in the same way that the UPC (bar code) does in the physical world (click
image to enlarge).
Further information:
Content Directions, Inc. (CDI): www.contentdirections.com
International DOI Foundation ("DOI" and "doi.org" are trademarks of the IDF): www.doi.org
Corporation for National Research Initiatives (CNRI): www.cnri.reston.va.us
[1] For a front-page New York Times article on granular publishing, see the boxed article "Books by the Chapter or Verse Arrive on the Internet This Fall," by Lisa Guernsey (New York Times, July 18, 2000, p. 1).
