Is "Semantics" the Next Big Thing?
Bill Roth's Blog |
July 25, 2007 3:01 PM
|
Comments (2)
As I was traveling across Asia and hanging out in waiting rooms, customs lines, etc., my mind turns to the future, since the present is so dull. In our business you always have to keep wondering "What is the next big thing?". The more I think about it, "Semantics" always seems to bubble up to the top. To be clear, Semantics is the study of meaning. But its much more than that. Everything old is new again Note that I did not say Semantics is the next New thing. In fact, The pursuit of "Semantic Technology" is by no means a new pursuit. The earliest research on semantics and computer science/linguistics research dates back to the late 1950s to a Noam Chomsky book entitled Syntactic Structures. An an endeavor, we have been seriously looking at ascribing meaning to objects and concepts since the 1958 arrival of the programming language LISP. So, clearly this is not a new effort. But all of the previous attempts at doing anything meaningful (no pun intended) did not bear practical fruit. from the 1950's to the early 1990's there was a substantial amount of research in Computer Science in these areas, especially machine learning. But something fundamental has changed. The rise of XML has given us a set of lexemes and syntax with which to scaffold upon to deal with semantics. In essence, we now are heading toward a future where computers have the ability to take away those small, boring decisions we have to make in every day business. The Search For Meaning Having a set of words and a syntax doesn't necessarily mean you can derive meaning. There has to be a common understand as well as a common hierarchy of knowledge, known as an ontology. Once you have this, you have the basic abilities to do simple reasoning based on meaning. Now, the Web has been a quantum leap in humanity's ability to gather and find knowledge. But, as Tim Berners-Lee noted in his seminal work The Semantic Web, "the World Wide Web has developed most rapidly as a medium of documents for people". Thus, it limits the Internet to merely sending text and images around. What we need is to augment the web with more information, so more inference can be done by systems, and not people. Why "Semantics" is Important to Business Now for the obvious question. Why is all this abstruse research relevant to business? To answer that, you have to go back to first principles. What is enterprise software all about? Automation. Taking people out of the loop. Improving cycle time. Improving responsiveness. So what good is semantics? Consider a workflow in a business. There are a series of steps in the workflow in which some of the steps are done by computers, and some of them are one by people. Most often, the ones done by people, are because some inference or decision has to be made on the incoming document or data. If we begin to augment the information on the web, more of the decisions, and more of the inference-ing, can be done by computers. This will lead to faster overall completion of workflow, by removing the off-line human element, where possible. This results in saving money and greater productivity, and this is why "semantics" is important to business. A Partial Bibliography on Semantics and the Semantic Web Feel free to post other links if you think they are relevant.
Comments
Comments are listed in date ascending order (oldest first) | Post Comment
-
Also posted as http://java.sys-con.com/read/407403.htm
Posted by: wgroth2 on July 27, 2007 at 8:53 AM
-
Hi Bill,
The concept of semantics is coming back under different names.
In one sense, semantics is coming back as Web 3.0(or Semantic Web 3.0);
Nova Spivack is a great proponent of this and his company Radar Networks, is one of them trying to enrich the html pages with an RDF based summary and assign more meaning to static pages. There is also another company named MetaWeb operating in the same space. I believe most of them are in the stealth mode. Those are the ones operating in the consumer space, and mostly come from AI background(most of these guys ae from Ray Kurzweil's Thinking Machines ;
In the IT space, semantics is coming back under the name of Complex Event Processing; I have done work on this in 2004, collaborating with researchers from IBM labs with their AMIT Project(Active Middleware Technology). The main goal here is to derive meaning so that things could be automated better.
Yeah, Everything old is new again :)
-Radha.
http://popuri.blogspot.com
Posted by: hiradha on July 30, 2007 at 11:22 AM
|
