Part 1: What Is It
(This is a three-parter. In this first post I explain what the semantic web is, or may become.)
If you’re not ready for it, the next big web evolution could put you right out of business.
If you read the business magazines, you’ve seen the phrase ‘semantic web‘. It may seem like a bunch of academic gobbledygook, but it’s not. The semantic web that’s slowly emerging will let customers more easily find you, compare you to others and make a decision.
The Web Now
Search engines go out, read web sites, and make their best guess as to what each page means. If one page has, say, 30 products on it, the search engine doesn’t differentiate between those products – they’re all part of that page.
You want to buy a running shoe. Go to Google and search for ‘running shoes’. You usually get a result that reads like this:
“I HAVE SHOES!”
“SO DO I.”
You click a link. If you find what you want, you stay and buy. Or you click ‘back’ and return to Google.
If you’re searching for an internet marketing consultant, then you type in ‘internet marketing consultant’, and see a result like this:
“I AM THE GREATEST CONSULTANT EVER”
“I WON’T RIP YOU OFF”
“I’M A VERY EXPERIENCED MARKETING CONSULTANT”
“NO HE’S NOT. HE’S EVIL! TALK TO ME INSTEAD! TRUST NO ONE ELSE!!!!!!!!”
The Semantic Web
Search engines go out and read web sites. But now, instead of guessing what’s what, they read information about the information, called metadata, and classify each ‘chunk’ based on that data. If one page has, say, 30 products on it, the search engine can separate each product into a discreet item and let you search them individually. So your result might be:
“I AM A SHOE!!!”
“An athletic shoe. With a nice sole and stuff. My price is $30 and you can order me online.”
“I AM A SHOE, TOO.”
“A running shoe. Designed for trail running. My price is $200 and you can order me online, too.”
These listings might be from one site, two sites, or five sites. You can read each one and decide what to do.
Revisiting our consulting search, you might see a result like this:
“In business for 10 years. Clients include X and Y. Located in Seattle. Has a blog you can read at Conversation Marketing.com.”
“In business for 11 years. Clients include X and Y. Located in NYC. Published in Fortune Magazine last month. Works with John Smith.”
The advantages are obvious:
- Search engine developers and others can build tools that automatically classify information about products and services.
- Consumers and researchers can more easily find what they need.
- Search engines are less likely to be confused by the difference between, say, ‘laptop’ and ‘notebook computer’ than they are right now. Folks get an accurate picture of your site, and they’re more likely to buy when they get there.
Best of all, you have greater control over how search engines classify and place you. If you do it properly. More about that in part 2.
Cool! Where Do I Find it?!
Unfortunately, nowhere yet.
There are little bits of the semantic web emerging. But developers, information architects and designers have to implement semantically useful data, first. Which they won’t do until someone creates a search engine that really relies on semantic data. Which won’t happen until developers, et al implement semantically useful data…
Sounds grim – it’s a chicken-and-egg problem with no chicken, and no egg.
Don’t despair just yet, though. Recent patent filings by Google show they’re ramping up for a major change in their engine that will build on and grow a semantic web.
Finally, there are already bits and pieces of the semantic web in use: RSS is one example. Microformats (less-known but increasingly popular) are another. Hakia attempts to extract semantic data from existing content, and does it pretty darned well. And sites like Technorati use community-generated tags to semantically organize their directories.
The semantic web is coming, whether you’re ready or not. Next time, I’ll talk about what you can do to stay ahead, and even take advantage.