Content labels another answer to the growing search problem?

I wrote yesterday that Google and others are trying to make the search problem more manageable by carving it into smaller chunks – be it vertical search or splitting the problem horizontally.

Another approach for cutting out the dross and getting quickly to the right pages is content labelling.

Content labels are machine readable metadata attached to sites that operate as an official stamp.  Verisign operates a content labelling system where sites that pass security standards are allowed to show a Verisign trustmark which can give people enough trust in a site to enter their credit card details.

Other prominent areas for content labelling are accessability – sites that are suitably set up for the visually impaired can carry an appropriate trustmark – and trustmarks indicating suitability for children and readiness for mobile.

Trustmarks/content labels help with the search problem if browsers or search engines can be set to only return search results which carry a certain content label.

Breaking that down, for the system to work the following elements are required:

  • An agreed standard
  • A certification authority
  • An effiicient process by which sites can get certified
  • A content label registry where trustmarks can be looked up and verified
  • Browsers which can read and check the trustmarks

Ireland/UK start-up Segala run by Paul Walsh is looking to deliver the last three parts of the equation.  They have a technology that enables efficient certicication of sites, a web service for trustmark look-up and verification and a Firefox plug-in that solves the browser part of the equation.

It is early days yet, but I like this because it is a solution to part of the search problem – it would be cool if a search on my mobile phone only returned sites that were certified as mobile ready – and in theory this could be extended in almost any direction.  Cites could be certified as “business” or “blogs” for example, which would really help weed out the unwanted from search results.  It might even be possible to include socially generated metadata in some way.

I also like it because it is a potentially huge play – it could be pennies on almost every website on the planet as a charge for certification – and there are now 45 billion of those.

But it all hangs on standards being agreed, and bitter experience tells us what that means – it will certainly take forever and it may never happen.  There are lots of encouraging signs though – with dotmobi pushing their mobileOK standard, the W3C getting in on the game and and ICRA and Ofcom are promoting content labels as a method of parental control. 

  • Thanks for the mention Nic! 🙂

    Regarding the first two bits that Segala isn’t covering 😉

    Content Labels is now moving onto a W3C full recommendation track. They will be proposed as a replacement for PICS – old recommendation still used by Internet Explorer for filtering content.

    The new XXX top level domain name soon to be given the green light will mandate all registered sites to use a Content Label.

    PEGI – has stated that it will use Content Labels for games rating.

    Keep an eye on http://contentlabel.org where we intend to suggest solutions for the great use cases you gave 🙂

  • Thanks for the mention Nic! 🙂

    Regarding the first two bits that Segala isn’t covering 😉

    Content Labels is now moving onto a W3C full recommendation track. They will be proposed as a replacement for PICS – old recommendation still used by Internet Explorer for filtering content.

    The new XXX top level domain name soon to be given the green light will mandate all registered sites to use a Content Label.

    PEGI – has stated that it will use Content Labels for games rating.

    Keep an eye on http://contentlabel.org where we intend to suggest solutions for the great use cases you gave 🙂

  • Forgot to mention. mobileOK referenced in your note – conformance claims about mobileOK will come in the form of a Content Label. So, it’s early days for sure, but I think 2007 is the year we’ll see them come into play as we get the ecosystem going…

  • Forgot to mention. mobileOK referenced in your note – conformance claims about mobileOK will come in the form of a Content Label. So, it’s early days for sure, but I think 2007 is the year we’ll see them come into play as we get the ecosystem going…

  • I call this area the “metadata wars” – but there are many ways to create metadata, the crux is owning the “best” site that has “good enough” metadata in the niche desired.

    Certification is a subject in itself, suffice to say I think the solution is not going to come from multiple certification authorities.

  • I call this area the “metadata wars” – but there are many ways to create metadata, the crux is owning the “best” site that has “good enough” metadata in the niche desired.

    Certification is a subject in itself, suffice to say I think the solution is not going to come from multiple certification authorities.

  • Alan – the good thing about Content Labels is that they are based on the Semantic method, RDF. This is the most sustainable form of data available – everything else is a tiny implementation of the Semantic Web.

    I think we’ll need a number of ‘Certification’ authorities who specialise in different areas. For example, I wouldn’t profess to specialise in security. However, that’s not to say we couldn’t live with just one ‘labelling’ authority 😉

  • Alan – the good thing about Content Labels is that they are based on the Semantic method, RDF. This is the most sustainable form of data available – everything else is a tiny implementation of the Semantic Web.

    I think we’ll need a number of ‘Certification’ authorities who specialise in different areas. For example, I wouldn’t profess to specialise in security. However, that’s not to say we couldn’t live with just one ‘labelling’ authority 😉

  • Ruby Akbar

    Most people ignore most poetry

  • RUAK0876ac

    Most people ignore most poetry