Google’s dominance in online search hasn’t stopped hundreds of startups from trying to build a better mousetrap. Each is trying a new twist on search: geography, crowdsourcing, tags, user annotations, learned hierarchies and timelines. With $20 billion spent on online advertising every year, a killer search application can make a lot of money.

But will new types of search catch on? A recent study of the Google Generation, conducted by University College London, found that “users make very little use of advanced search facilities, assuming that search engines ‘understand’ their queries.” Many of today’s Internet users still don’t know how to use a search engine, preferring instead to type a domain name into the search box (which is why Yahoo is a top search on Google and vice-versa.) The reverse, known as type-in traffic, involves typing a search topic into the address bar to find results.

Alistair Croll‘s opinion

If the flurry of search startups can tie into the social graph of Google, Facebook and others without biting the hands that feed them, then they have a chance of succeeding. But if they’re betting their business on changing the way people search, they have a lot of work ahead of them.

I totally agree with Alistair. We don’t need more search engines. What we need is for the current search engines to do all they can to enable more relevance and trust. Battling against Google is mad to say the least. This is why Segala has been working on Content Labels for the past two and a half years. This blog is attracting new readers all the time so I won’t assume you know what I’m talking about. So, I’ve published below, an old post which explains what Content Labels are, how they work and importantly, why I think they will see mass adoption within the next 24 months.

How to enable current search engines provide more trust

Before I explain how Content Labels enable more reliable and relevant search results, I will first use an analogy to help explain in plain English, what a Content Label is. This will help you to visualise yourself making use of a label, whether you are an end user, developer, Web site owner, search provider, browser vendor or trustmark provider.

Think of Content Labels on the back of food packaging

Jackie goes to the supermarket to buy the weekly shopping. Amongst other things, her shopping list includes a tin of baked beans. Most of the time, Jackie is happy to buy Heinz without hesitation.

Adrian goes to the supermarket with a similar shopping list. He however, has dietary requirements to consider. Adrian likes to check the Content Label on the back of the beans tin (irrespective of how much he trusts the brand) to ensure he’s happy with the contents before making a purchase.

Joanna is pregnant and although she should be taking it easy, decides to do the weekly shopping. She’s aware that eating any form of cream cheese is dangerous for the unborn. To ensure she doesn’t endanger her baby’s health, she checks the Content Labels on items that may contain cream cheese before making her purchasing decisions.

The moral of the story

People have different needs, some will only buy products that are low in fat, low in calories, high in protein or have some other dietary consideration.

Some brands will do more than provide a Content Label to promote the contents of their products. Brands such as Weight Watchers, for example, will promote specific aspects of their labels as a unique selling point to differentiate them from competitors. This allows buyers to look specifically for Weight Watchers without searching for any other brand to make a purchase, even though other brands may be trusted. Other brands are simply not ‘relevant’ (at this time).

So, there are numerous reasons why people check the Content Label on food packaging. The main point to highlight is that we are all different, with different needs and preferences when searching for and buying products.

If it wasn’t for the Content Labels on the back of food packaging, we would end up having to search and buy lots of products in order to taste and discover the ones that meet our dietary requirements. Some people may even die if they eat something they’re allergic to (such a nuts) – this may seem like an exaggeration, but the analogy does apply to the Web – consider a user who buys medication online.

If you’re the type of person that doesn’t care to read Content Labels on packaging, at least keep in mind that a lot of people do.

The problem today

The only information search engines provide today is the title and description of each Web site, nothing more. How do you know which Web sites are trustworthy?

A range of trustmark (visual badges) schemes exist around the world. In each case, a Web site is awarded the right to display a seal of approval or trustmark following a review by an independent authority or internal assessor. Trustmarks normally hyperlink to a certificate page provided by the labeling scheme provider to authenticate its claims.

However, trustmarks are invisible to search engines which limit their value. A user must already be on a Web site before they know it meets a labeling provider’s claims. Moreover, the logo is often displayed on just a single page and therefore invisible to visitors who are viewing other pages on the site.

The solution

Taking the food analogy to the Web, wouldn’t it be beneficial if Web site owners could provide more information about their content before users hit their site? In other words, allowing users to read specific information about the suitability of their content from search results. Well, this is now possible using Web Content Labels.

• It’s now possible for site owners to provide more information about their content from search results.
• It’s now possible for search engines to provide more reliable and relevant search results.
• It’s now possible for users to read Content Labels and make informed decisions about the suitability of content before entering Web sites.

Content Labels are files used to classify content in the form of metadata. The information contained within a Content Label is not generally found on the Web site, unless it’s in the form or a Trustmark. Content Labels not only contain information about the content, they also contain information such as the company name, asserter, creation date and expiration date.

The type of information contained in a Content Label very much depends on what the site owner wants to make claims about. It’s important to note that Content Labels have a very specific use. They’re used to make assertions about conformance to industry standards and codes of conduct. They’re not used to make assertions that are ambiguous. Although it’s debatable that most standards and codes of conduct are open to interpretation.

In concept, Content Labels act in a similar fashion to SSL Certificates. SSL Certificates are used so browsers can tell if a site is secure for credit card transactions. However, Content Labels are much more flexible. They can be used to make claims about a specific URI without the restriction of having to make a claim about an entire domain.

Content Labels are a lower cost solution than SSL Certificates as sold by companies such as VeriSign and GeoTrust. Oh wait, VeriSign bought GeoTrust which was their biggest competitor, so that provides them with a monopoly using proprietary technology. This is something we intend to disrupt!

We’re also helping to build other elements of the ecosystem. We have instigation the creation of ContentLabel.org. Contentlabel.org facilitates the creation and harmonization of new codes of conduct for which Segala will create new Content Labels. The idea is for any organisation or association worldwide to create a code of conduct and then have it reviewed in an open forum, formalised, endorsed and promoted.

It’s also possible for organisations and associations that already have a code of conduct (even if they have a trustmark) to have it recognised by industry and promoted to encourage mass adoption.

We intend to start with a new code of conduct for blogs as it appears to go down well with the bloggers we’ve spoken to. Anyone can participate so get in touch if you would like to get involved.

Share it

I stated recently that I wasn’t go to Chair or speak at conferences this year as I had my fill of them last year. I’ve already had to decline 2 kind offers to Chair very good conferences. So I’m being very picky about the events I attend this year. If I’m to Chair or speak, it must be directly related to Segala’s business, be beneficial to the community (i.e. it’s no more than £40 to attend and not the typical £750+), or be in a sunny location Don’t get me wrong, I love to be asked. I just don’t have the time these days.

The Irish Web Technology Conference is my first speaking slot of the year.

What I’m going to talk about

I’m going to deliver a presentation on how Content Labels (AKA POWDER) will enable more trust on the Web. I’ll have with me, a tin of Heinze Baked Beans to help me demonstrate how Content Labels help consumers make better informed decisions about the suitability of products before making a purchase.

Sound interesting? Well, why not come along. As with every conference, you’re likely to get most value by hanging out in the hallway chatting to others and with around 500 delegates expected, you’re likely to meet some very interesting people. I might give away a prize as a means of keeping everyone awake. I’ve done this before and it seems to work well.

Oh by the way, all of this Content Labels malarkey is based on the Semantic Web. Don’t tell anyone though as they might assume it’s academic, boring or a nice concept.

About the conference

WHY?

This event’s aim is to showcase the sheer brilliance of the Irish Web Industry. It’s a four day, 10 track, 50 session web extravaganza.

WHAT?

The inaugural IWTC 2008 organised by IrishDev.com in association with the Dublin Java User Group and First Port Jobs, is the first of seven major technology conference in our Series 2008.

IWTC 2008 is an event by the Irish Java community for the Irish Java community and will bring many experts and internationally renowned speakers together to discuss latest trends and emerging technologies.

With almost 50 expert technology sessions to choose from, covering a breadth of Web technologies, this is yet another tech conference you can’t afford to miss!

When?

Begins Tuesday 26th February, Ends Friday 29th.

Where?

Cineworld Complex Parnell Street, Dublin 1

Share it

One of the standards I’d like Microsoft to dump however, is PICS (Platform for Internet Content Selection). PICS was one of the very first W3C recommendations and the first to be adopted by Internet Explorer during the mid 90’s. It’s still in use today by the Content Advisor functionality (see screen shot below) to help parents protect children from inappropriate content.

PICS however, is inaccurate and it’s certainly not flexible enough to be meaningful in any way. This means that the filtering tool in IE is obsolete. It’s not Microsoft’s fault though, there hasn’t been an alternative until now.

What’s the alternative?

PICS will soon be formally replaced by a new method of classifying content called Content Labels in approximately two to four months. Before I move on and in case you haven’t read this blog before now, a Content Label put simply, is a file that contains Metadata on steroids. It’s this Metadata that would enable IE to learn more information about Web sites before presenting search results to end users.

As we’ve demonstrated with a Firefox extension in the screen shot below, IE could place a tiny icon beside each search result. This would tell users whether each Web site owner was making an assertion about the suitability of their content or not. As you can see from the screen shot, we’ve placed a red x beside search results where the Web site owners are not making an assertion of any kind.

The red x could just as easily be a question mark, or an exclamation mark. The search result with a green tick represents a Web site which has been independently verified by a trusted third party. The screen shot doesn’t display any search results with an amber tick but this could reflect Web sites that are making an assertion which hasn’t necessarily been verified by anyone.

Upon clicking on each icon the user is presented with a certificate, containing more information about the suitability of the content on that particular site. The certificate in the screen shot below isn’t particularly meaningful right now, but it could have a list of icons on the left side with a description of what each one represents to the right. Users could then select each icon to find out more about the actual assertions made by the site owner.

Based on the same Metadata, IE could build a better version of content advisor so users could filter Web sites based on more accurate and meaningful personal preferences. That’s what you call personalised search based on trust, relevance and safety.

Tim Berners-Lee, inventor of the Web, Ivan Herman, Semantic Web Lead and others at the W3C, felt that the term Content Label would give the perception that we wanted to police the Internet. So, they asked that the name be changed to POWDER (Protocol for Web Description Resources). Don’t ask and I won’t confuse you any further.

I must admit, I was against their request for a name change but after delivering a few presentations, I quickly realised that some people, irrespective of how many times you say otherwise, think Content Labels is about policing the Web. So we decided to agree to the name change for the sake of keeping the peace and moving the project forward.

Whilst we refer to our method of classifying content in the W3C as POWDER, we will always refer to it as Content Labels outside of the W3C, as we find that it’s easier for our audience to relate to labels. I can’t even remember what POWDER stands for most of the time

So, in summary, POWDER and Content Labels is the same thing.

Why IE should read Content Labels to enable more trusted search results

1. Content Labels will replace PICS as a ratified standard in approximately two to four months, thereby making IE’s method of filtering obsolete as I said earlier.
2. IE can highlight Web sites that make assertions about conformance to industry standards and best practices. Type of assertions include accessibility, adult, privacy, copyright, eco-friendly, medical, marketing and more. PICS is only capable of making one type of assertion.
3. PICS and every other implementation of trust on the Web today, including SSL Certificates for security, are extremely limited in functionality. They only permit you to make an assertion about an entire domain. Content Labels on the other hand, are so flexible that it’s possible to make an assertion about an entire domain, or about specific Web pages.
4. Content Labels are based on the Semantic Web and therefore future proofing the Web for better content discovery.

How it works

Without getting into the technical detail, the browser would identify a Web site with a Content Label by locating a link tag in the header, just as it does with the title and description tags. It looks something like

< link rel=”meta” xhref=”http://www.segala.com/labels/tcuk_label_001.rdf” type=”application/rdf+xml” title=”Segala label” /

The browser then reads the contents of the Content Label by following the link above. Self labelled sites can be stored by site owners. Sites that have been independently verified reside on the trusted providers’ server.

Read more about Content Labels, the use cases and how mass adoption can be achieved.

Share it