Idea For A Better Search Engine Algorithm?

[DaveSawers]

I've been thinking about how to build a better search engine algorithm.

The aim as I see it would be to base ranking on content alone and to rank sites with original content higher than those who just steal or copy it.

To make it work would need a copyright database to which new work could be submitted before publication and which would have algorithms to detect plagiarism as well as outright theft. Sites could still refer to other sites and quote from them but if they didn't include proper attribution, would not get their pages indexed.

The advantages would be:

That sites with original content would always rank higher than those who just aggregate or steal from other sites.

Directories, Adsense traps, link farms and other internet garbage could be made to disappear from search engine results.

Incoming links from other sites become irrelevant so spammer bots that plague blogs and forums with useless comments are no longer effective and hopefully, cheating the system becomes much harder.

The main disadvantage would be the building and maintenance of the copyright databases and the handling of copyright disputes.

More information is contained in my latest blog topic.

Do you think this might work? Am i just crazy? Is this an original idea even?

[DomainMoon.com]

how can you authenticate that you are the real developer of the content ?

[DaveSawers]

Originally Posted by DomainMoon.com

how can you authenticate that you are the real developer of the content ?

The same way copyright is decided now. By being the first to publish it.

[stymiee]

Google can already do this with their algorithm. That's how they filter duplicate content. The original get s credit for the work and the copiers go to the supplemental index. No database needed.

[Chris]

bttt

[Feydakin]

I can't see the bulk of people going through the extra effort of submitting to your database first and then publishing on their own website.. It might work for niche markets, but for the generic 'web'??

Add to this the very well done scripts runinng out there.. The ones that aggregate data and create relatively unique content for adsense.. Turn this loose, add the step where it submits to the copyright database, then publish.. Unique content, but still junk designed simply to get you there and then click out on a link that they get paid for..

Unique on page content doesn't seem to be enough to sort the web these days..

[kgun]

Very interesting subject. It is not easy.

Scroll down to

• Grid computing.
• Then Oracle Secure Enterprise Search.

You may be able to get some ideas from those links.


You should also, by using digital certificates, or a good authentication system, be able to make a ranking system that is better than what we see online today. It should be very difficult to vote more than once.

Remember: Quality is often better than quantity.