Tuesday, August 14, 2007

Extract Meaning from Text, as a Webservice!

Hublog has some has some sweet APIs under its hat.

The one of most interest is the Clearforest web service. If you want to be able to extract people, places, organizations, cities, and a whole lot more from a bunch of text, this is worth it to you.

I spent a bit of the evening playing with text extracts from Wikipedia, local government, and all sorts.

It's quite nifty; but doesn't quite pick up everything I'd like - ie, I tried it with a fragment on Stamps from wikipedia.
Postage stamps of Ireland are the postage stamps issued by the postal authority of the independent Irish state. Ireland was part of the United Kingdom of Great Britain and Ireland when the world's first postage stamps were issued in 1840. These stamps, and all subsequent British issues, were used in Ireland until the new Irish Government assumed power in 1922.

It picked out everything about the countries, governments, etc, but couldn't tell me about Stamps, or about years.

Give it a shot and see what you can do!

2 comments:

Anonymous said...

Thanks for taking note of our semantic web service! I'd also like to point you to our Firefox extension @ http://gnosis.clearforest.com. Gnosis uses the same web service to "mark up" people, places, companies, etc on the web page you're currently viewing and automagically launch searches of Google, Technorati and other sites.

Also, stay tuned for some significant functionality enhancements to SWS in the coming months. Much deeper extraction, higher performance and generally cooler.

ClearForest

Dan said...

Yeah, I've had a play with it; and it didn't quite push the right buttons for me.

Keep up the good work though on the parser and webservice