What is the 'semantic web'?
I keep seeing people talk about the ‘semantic web’ and have no idea what it’s all about. So, in an effort to try and understand it, I thought I’d catalogue my journey in the hope that I might understand it by the time I finish.
Tim Berners-Lee (inventor/creator of the WWW) classifies the semantic web as “a web of data that can be processed directly and indirectly by machines” (source: Altova). That doesn’t really help me out though.
I first heard about the semantic web via a few entries on Twitter, and then a newsletter from the eLearning Magazine popped in to my inbox all about “The Semantic Web and E-learning“. This article is slightly better in its explanation, saying that;
“the Semantic Web is a collection of standards, data structures, and software” and that “semantic technologies seek to define the framework and method of communication between systems.”
That’s slightly better … it’s all about standardisation.
The difficulty I have found in trying to explain the semantic web is that it is too big for one simple explanation (if there is one way to explain it). Infomesh helped me to understand it a little better with their introductory article, where they say;
“… just think of information about local sports events, weather information, plane times, Major League Baseball statistics, and television guides… all of this information is presented by numerous (web) sites, but all in HTML.”
The semantic web will bring the content (not background code, but purely the content/text/information) on the page(s) to a “repurposable form” that can be used and re-used across all platforms, all locations, etc … kind of like one big database that we can all dip into (or contribute to) for whatever re-usable content we need. I think.
Hell, it’ll take a whole lot of reading to fully grasp what it is and where it is going, so if you have information or links that will help me on my way, please let me know in the comment box below.
Hi David,
I too am trying to learn the Semantic Web. I found reading about RDF (http://en.wikipedia.org/wiki/Resource_Description_Framework) really helped conceptualise the underling technical framework for the Semantic Web. Really, really, really, tough going wading through the documentation, but I found information about the FOAF project (http://esw.w3.org/topic/FrontPage) very useful. This is a small and relatively basic implementation of what the Semantic Web (using RDF) is about.
If I come across any other interesting stuff I’ll let you know :)
Hi David,
I’m still new at this too, but I think I can help with a conceptual framework.
The semantic web is supposed to solve a problem. That problem is this: Computers don’t speak English. Moreso: humans don’t speak computer. So, in any search-related activity, we are effectively engaging in translation – from English (or any other human language, also called ‘natural language’) into computer and back again. That’s messy & inefficient.
Semantic means ‘meaning’. If I type ‘horse’ into google, I (as a human) know it means ‘nice hoofed mammal you can ride on’. I know the string ‘horse’ is a actually a word that represents a concept, and the concept relates, in one way or another, to other concepts that can be represented by strings – ‘mare’, ‘gelding’, ‘foal’, even ‘saddle’. They’re not synonyms, but they ARE related.
As far as computer language is concerned, the string ‘horse’ has as much relationship with the string ‘mare’ as it does the string ‘antidisestablishmentarianism’. That’s the problem the semantic web is trying to fix.
So: the semantic web is a method for creating computer-readable relationships between natural-language concepts. It’s about coaxing computers into speaking just a little natural language.
A good overview & much-needed criticism can be found here: http://bexhuff.com/2009/01/the-semantic-web-impossible-in-theory-impractical-in-reality
ps. I’m coelacanthro on twitter
At TED 2009 in Long Beach, Sir Tim Berners-Lee modified his pitch about the next generation web, i.e. Semantic Web, to something a little bit simpler to grasp — LINKED DATA. The Web, as he originally conceptualized, is properly referred to as LINKED DOCUMENTS. :D
Hi David,
Good post – really summarize well.
If you’re interested in the semantic web, you can check out – http://www.urlclassifier.com for extracting main topics from URLs – utilizing some of ContextIn Semantic Advertising algorithms
Thanks for being an eLearn Magazine reader and congrats on the baby!
Thanks Lisa; loving fatherhood and trying hard to make time for blogging and sleep, anything else is a bonus!!