The Semantic Web has enormous potential to change the way we receive, understand and use information. The Web as we know it today connects pages of information one dimension at a time to each other based on some simple things you ask it to perform (e.g. keywords like “dog” & “food”). Of course you get some pages that talk about dog food. But many others that simply happen to have the words dog and food somewhere on the page yet talk about all kinds of things other than “dog food.” A Semantic Web makes sure the concept of dog food is present first, and then identifies other facts, experts, types, uses, recipes, ingredients, etc., about dog food. A Semantic Web is smart in that it presents a better set of results, in context and is ready to solve problems, answer questions directly, infer, resolve, discover and analyze in ways that the current web was never designed to do.
Still, the Semantic Web is still in its infancy. For all its potential, the Semantic Web still has a long way to go. Why? Education, exposure, and evangelism are part of the reason. Actual experimentation and perfection of technology that drives the Semantic Web are two key additional reasons. These take time, sometimes many years. The VCR, telephone and cell phone each took nearly 10 years to reach a 20 percent use mark among the general population. In comparison to these technologies’ adoption rates, the Semantic Web is still in its early stages of use. The 20 percent penetration mark is often looked at as the first inflection point in the often cited S-shaped technology adoption curve where the 21 percent through 80 percent penetration mark moves rapidly over time.
Standards help drive technology adoption but aren’t the end-all, be-all. Besides the factors above, standards will help settle the uneasy, risky feeling among potential users that are teetering on placing a bet on burgeoning Semantic technology. Standards accelerate the adoption of technology since users know it will not become obsolete or be eclipsed by something else in short order. When all technology makers agree and adhere to a standard, it clears the air and assures users that the technology is here to stay. We have good standards now with the Semantic Web; standards that are setting the stage for acceleration on the S-shaped technology curve.
Once standards are established and accepted, applications must come next. Hands-on experience with well-designed and useful Semantic applications is key to spurring adoption and generating new ideas. Thus a virtuous cycle of application experience and application generation supports the adoption of the Semantic Web beyond standards.
Semantic Web technology must be able to read and reason over content like we do, but faster. Once standards are met we face a choice. How will we use those standards to make the Web Semantic? There are two choices: 1) Mark up every page by hand by reading, deciding what a subject, predicate and object are and then add URI’s to describe them, or 2) Use technology that can do this for us. Amazingly enough, even some of the largest technology companies expect the first choice, the hand mark-up method, which is both time and labor intensive. Do the math and you can easily conclude that the mark-up method alone would add 30-40 percent to the workload of every IT professional around the world.
An evolving two-fold approach to automating the mark-up of the Semantic Web. There seems to be little debate right now about which approach, linguistical or statistical, is superior in solving the problem described above. Nor is there any evidence of work being done to see how both approaches could be used to support each other and fill-in where the other falls short. Still, this dual approach may hold great promise in relation to fully automating the Semantic Web.
Vision and guts are rare but these are what will and are already driving initial use of the Semantic Web. The Semantic Web is something that is new, different and has not been tried before. There is risk in attempting, and sometimes failing, to build and use a Semantic Web application. But risk often comes with great reward. At Expert System, over the course of several hundred deployments of Semantic Web applications, our customers tell us they get back $400 for every dollar they invest. ROI has and always will clear the way for much more use, experimentation and success.
While there are many issues unfolding in relation to the Semantic Web, the reality is that we’re well on our way to finding solutions for many of the current problems mentioned above. As we approach 2011, we look forward to establishing a Semantic Web that is as ubiquitous as the Web is today.