VentureBeat - Latest Comments in Information pollution: Can semantic search save the day?

Re: Information pollution: Can semantic search save the day?

mortgagevladim — Thu, 11 Oct 2007 13:17:32 -0000

Hi
I've found thesite with the searching system
The site has domain name
I've read many corresnondences
I want to open a new theme
It will be interesting
By by,everybody
mortgage

Re: Information pollution: Can semantic search save the day?

Courtney Benson — Wed, 08 Aug 2007 12:04:09 -0000

Riza -

Sounds like your on to a much needed service offering. When will it be ready for prime time?

Re: Information pollution: Can semantic search save the day?

nick — Tue, 07 Aug 2007 20:10:31 -0000

Nice piece. I agree with many of your points.

"Mega information clutter" really changes the data management game in many deep and meaningful ways.

Round these parts, we call the clutter "data superabundance"...

Re: Information pollution: Can semantic search save the day?

John Nagle — Tue, 07 Aug 2007 20:03:10 -0000

This has been tried before. Cycorp has been claiming that they'd have something like this working in two years since about 1990. What they have so far can be tried at "http://game.cyc.com". You won't be impressed.

We've discovered that kicking the junk off the web is a solveable problem. Try our "http://www.sitetruth..com". This is an automated due diligence system tied to a search engine. It tries to find out who's behind a web site, and lowers the search positioning based on that info. All those "domaining" sites, link farms, referrer pages, and similar junk just drop out. Remember "on the Internet, no one knows if you're a dog?" We don't let the dogs in.

Re: Information pollution: Can semantic search save the day?

Ken Ewell — Tue, 07 Aug 2007 16:15:28 -0000

What an informative article. I like the SST moniker too!

One thing about the semantic web though: Information polluters would have to go the extra step of creating a namespace and rdf tuples for their pages; not so with SST. So it is likely that information pollution would be less of a problem for semantic web applications.

Data cleansing is not a new problem and this is probably why closely engineered and monitored content, though costly, is not a such bad idea. Where do you think the saying "garbage in, garbage out" came from.

One inherent problem is that search engines present links that are used to drive traffic to "the right place". People that want to scam you will fight to be that right place. The scam is to appear real, authoritative and accurate. Just like you said in the end, Riza, how can you filter “There are WMD in Iraq”?