Imagine a scenario where you are preparing to make travel plans, for business or pleasure. If you had a crystal ball and could predict the probability of arriving at a destination on time, you could manage your time much better than you do now. Unfortunately, that is not the case today, despite the tremendous advances made in weather predictions, traffic delays at airports and road congestions due to construction etc.,

Put in simple words, such a scenario would represent a better world where an event takes place just-in-time, just-as-predicted and just-about-right

However, humanity has now at its finger-tips the tools to predict, to a high degree of accuracy, the natural disasters and weather patterns. We have several disparate tools to predict events. However, we do not yet have a composite tool to put it all together.

Consider, for example, the question posed by this curious user about lack of mash-ups for Air Traffic delays:

http://paul.kedrosky.com/archives/2007/10/03/air_traffic_mas.html 

Orbitz has made a note-worthy collaborative effort to provide such a mash-up, which relies on user updates. See below:

http://updates.orbitz.com/

Absent reliable user updates, however, this tool lacks the predictability and reliability.

The following weather mash-up sites provide information about weather across several geographies:

http://www.accuweather.com/maps-satellite.asp?partner=accuweather&traveler=0&site=CO_&type=ei&anim=1&large=0

http://www.weatherbonk.com/weather/summary.jsp?where=usa&_id=&_className=weathermaps.weather.user.WeatherRequest

What is missing is a mash-up or tool that provides semantics to all these inanimate data sets. For example, a web-site that accepts a user’s question: 

"I am traveling to Bentonville, AR on 11/27/2007, leaving from Cincinnati, OH at 13:00 hours, flying non-stop using General Aviation facilities in my Cessna Citation. What are my chances of arriving at my destination by 17:00 hours?"

We have the current technology to answer the above question in a more or less accurate manner. However, we do not yet have the ability to analyze the question, interrogate various mash-up tools to gather answers to sub-sets of the question and stitch it all together.

This is where we transition from inanimate Web 2.0 mash-ups to intelligent systems that give semantic meaning to the data provided by such mash-ups.

Given the finite nature of relevant questions in any particular domain, air traffic and travel in this case, it is possible to parse the questions into sub-sets and get answers to those questions using existing tools (mash-up) and mash together the results of those mash-ups.

Take for example. OWL Specification from W3C:

http://www.w3.org/2004/OWL/

and comments on how OWL is relevant:

http://www.xfront.com/why-use-owl.html

The example above describes how the computer can arrive at a logical conclusion that is not explicitly stated in the data presented to it. Similar conclusions can be drawn on predictability of air travel using disparate sets of data relating to weather, traffic delays and airport information.

That day is not too far off.

Sastry Dhara