Connecting the Dots

What is commonly understood about the Semantic Web is not what I care to write about. I have a long professional history of looking to the future of business and how things come together in unique and challenging ways. In other words how stuff turns out that most people did not anticipate and thus did not factor into their decision making - sometimes with terrible consequences. This is no doubt true also with the Semantic Web.

The more interesting question to me is what is it that can be done with the Semantic Web beyond what is talked about today or even several years from now.  If we are to continue building the blocks of the next wave of the web we have to think forward enough in how it can be used so we are doing the right things now. 

But lets start with what is thought of as the Semantic Web today.  It is intended to be a better directory.  That's it - at least for a start.  A Semantic Web is smarter than todays web because what the network understands the web page to be about is richer and with less ambiguity.  Lets take the case of a sentence like the following;

"John Ford owns a Ford Escape and loves to escape on the weekends to the mountains."

Todays web will serve up pages about a person named Ford and a car named Ford if I use that keyword.  But what if I am interested in only the person and not the car?  I have to cull through and waste time finding the page I really want.  Or if I am interested in the Ford Escape car and use the keyword escape I get a page about escaping to the mountains but not a page predominately about cars.  Examples like this abound and you know this to be true since we all use web search nearly everyday. 

Nova Spivak of Radar Networks does a nice job of explaining this quandary picking on Google to do so.  He shows how the web has evolved every 10 years or so as the amount of information has increased the tools we use must be reinvented to add new productivity.  We are now at the inflection point of keyword search technology where the problem described above drives productivity down as the amount of information increases further. 

The social networking web or web 2.0 gives us new levels of connections between information using tagging and so is an improvement in terms of relationships among people.  The Semantic Web uses true understanding of concepts to take this a step further and forms relationships between all objects represented on the web.   

So what would this mean for our sample sentence?  The Semantic Web understands that John owns a car named Escape and that its John who loves to escape and not the Ford car that loves to escape.  This is no simple task.  It involves understanding word forms, parts of speech, logic and intended use of words with more than one definition.  It also involves being able to do this for each sentence, among sentences, across paragraphs and whole documents. 

But it is here that the story begins to get interesting.  If the Semantic Web can play connect the dots within and between all the objects in a single document it means this can be done between documents as well.  And if you can do that a whole world opens up.

Everything about the environment in which businesses must operate have some degree of record about how that environment is changing.  And not necessarily after a full change has occurred do we learn about it.  On the contrary, if we can imagine what to look for we find that the incremental steps that lead to change become apparent and in many cases are a matter of public record.  In other words they are on the web. 

Think about a how the government talks about, writes about, holds hearings on the regulations it intends to change before it is voted on.  Think about the required paperwork that must be filed with police, fire, telecommunications, sewer, environmental and other service providers before the first shovelful of dirt is dug for a new factory.  Think about what patents tell us, what ads for new employees tell us, what financial reports tell us. 

Collecting enough of these, properly processing enough of these show us the footprints in the sand.  Not only of where things have been but where they are going.  The Semantic Web holds the promise of faster alerts for better decisions but also accurate predictions by increasing the degrees of freedom to act. 

Lets say you are a company planning to launch a new product.  The probability of the product being launched on a specific date 6 months into the future is far lower than the probability 3 months into the future.  Why?  Becuase far more things are going to change and have yet to be completed at 6 months out than 3 months out.   So the probability of knowing when the launch will happen increases over time. 

But it is also true that as a competitor your degrees of freedom to act gone down as time progresses.  After a competitor chooses an ad firm, or locks in a supplier with a contract the option for you to hire the ad firm or use the supplier is blocked. 

So the trick is to make a prediction about a competitors product launch at a time that gives you enough degrees of freedom to take effective action.  Afterall, if you learn of the product launch after it is made public there is little you can do about stopping or delaying it. 

The Semantic Web holds the promise of allowing companies to model, test and predict when a company will launch the new product before it happens.  Now that is a use of the Semantic Web that is killer.

The Semantic Web gets us a ground breaking ability to improve decisions with timely information, to increase options, to avoid common pitfalls, to preserve jobs, and to increase profitability.  Companies that embrace the Semantic Web will reap these windfalls.  Those who do not face being outmaneuvered and outsmarted.