In a recent blog post, Tim Berners-Lee, Director of the World Wide Web Consortium, wrote about a new paradigm. In the early days of the internet it was the computers and the wires that were important; more recently it was the documents that were the important things, the centres of focus and attention. Now, and in the future, “it’s the things they are about which are important.” Tim refers to this as the Graph, to distinguish it from the Web, and also to distinguish it from social network sites that contain a subset of a representation of a Graph, but are not the Graph. Tim also identifies the Graph with the Semantic Web: being a way to express the relationships between things sitting above the level of documents.

"It’s not the documents, it is the things they are about which are important."
Tim Berners-Lee

"Cashmere is more addictive than cocaine."
Richard Lawn, ColourMart

In a recent blog post, Tim Berners-Lee, Director of the World Wide Web Consortium, wrote about a new paradigm. In the early days of the internet it was the computers and the wires that were important; more recently it was the documents that were the important things, the centres of focus and attention. Now, and in the future, “it’s the things they are about which are important.” Tim refers to this as the Graph, to distinguish it from the Web, and also to distinguish it from social network sites that contain a subset of a representation of a Graph, but are not the Graph. Tim also identifies the Graph with the Semantic Web: being a way to express the relationships between things sitting above the level of documents.

It’s things that are important. Things like cashmere, lovely, lovely cashmere. Like the very addictive cashmere sold by Richard Lawn in his ColourMart store. Richard distributes free sample packs to potential customers, 25% of whom will go on to place an order. That’s a better response rate than first time cocaine use. Richard has equally impressive statistics for two elements of ColourMart’s on-line presence. In his Ebay store, Richard has received 100% positive feedback with a score of 3,312. Meanwhile, his Yahoo Group of 1000 members contains over 10,000 messages, the overwhelming majority being from satisfied customers sharing their skills, experiences and photos of the garments they’ve created with Richard’s yarns.

The web applications that Richard uses are not semantically rich. Although it has other features, the ColourMart Yahoo Group is used mainly as a mailing list, while the html markup of eBay’s feedback page is purely decorative. Even so, there is still a Graph here. Real users are talking about real experiences with real products. The Graph exists, irrespective of the servers or the wires or the documents or even the semantically encoded representation of itself.

With the simple use cases that

* A customer can rate product and service delivery
* A buyer can read ratings
* Customers can share experiences

ColourMart demonstrates that, in the right circumstances, applications with these features will be used by customers and will clearly communicate to current and prospective buyers the quality of products and services they can expect.

Richard has shown, concretely, that it is possible and beneficial to form and engage an on-line community of prospective and returning customers. Tim has described a paradigm and a framework for structuring that engagement. But how will we get there from here? Six potential steps along the way are:

1. Companies web presences will extend beyond their own web sites.

Media sharing sites such as Slideshare, YouTube and Flickr are becoming more relevant to businesses both as a way of hosting their own media files (where allowed) and as a means of connecting with their fans and detractors. In the meantime, the sites themselves are extending their contribution to the Graph with tagging, grouping, friending and geo-location features.

Gabriola Sea Kayaking’s Matt Bowes, for example, uses Flickr to store annotated and geo-tagged images of the beautiful locations he visits on the trips he runs, sharing them with colleagues and clients.

Meanwhile the application sites are opening up with feeds, apis, gadgets and widgets that allow content from a person or a group, with a tag or in a location to be published to a social networking site, a blog or a portal application. The content can therefore be shared with the Graph.

2. Small business use of web applications and social networking applications will increase.

Small businesses used to build web sites with manually crafted static web pages. Now the sites have become dynamic, with the pages being held in databases and served through blogs, shopping carts or content management systems. Often these are installed by design or technical consultants on overstretched shared hosting servers or under-utilised dedicated servers. Increasingly, small companies are choosing alternatives to the technical risks and maintenance overheads of traditional hosting, opting to use a web application provider to replace or supplement their traditional web site. Others build their presence within social networking frameworks because of a good feature match, or because the network contains their target audience.

Love’s Fish Restaurant of Brighton, England has used standard photo, review and event components to build their Facebook Page. Even though the site has only been up a month, they already have several favourable reviews from satisfied customers and a growing number of fans who display recommendation badges on their own profiles.

On a larger scale, the New York Times’ Facebook Page, with over 4,000 fans, contains custom applications that share the most emailed stories of the day and run news quizzes.

On Ning, Kirstin Sloan, a dancer with New York City Ballet, set up The (Inter) Mission community which provides its 600 members with a private space where individuals with a vested interest in the future of dance can network.

These hosted systems are already making connections that are important to the Graph. Ning, by default, generates applications that link social media into Facebook. Ning also acts as a container for OpenSocial applications, which means that a member of The (Inter) Mission can use the same applications they access in Orkut, Linked-In, Hi-5 etc.

In the long term, hosted applications can also make for a faster shift to Semantic Web technologies. With the data structures already in place, the addition of RDF, FOAF or microformat components would, in many instances, make possible the viral spreading of Semantic Web features. With the release of Facebook and OpenSocial apis, this can be accomplished by any interested developer, irrespective of the platform strategy.

3. Data about the products, services, locations, structure and culture of small businesses will become better organized and increasingly re-usable.

Large, hosted, editable public databases are becoming available, for free.
Google Base already contains information about 500 million products, 47,000 company profiles and 23,000 businesses for sale. In addition, in the US, Google is actually paying a $10 bounty for adding validated information about local businesses to improve the quality of its mapping database.

FreeBase is storing highly structured, collectively edited information about companies and their products, services, locations and internal structure. These systems again provide feeds and apis, allowing the data to be queried, extracted and transformed to fit into the Graph.

4. Browsers and web applications will become more semantically aware.

Browsers like Firefox and Flock are already implementing Semantic Web and Social Networking features either as plugins or as core features of the browser. With a properly configured browser and web page, it is already possible to perform actions on contacts, events, locations and semantic tags embedded in web pages. These features will grow as semantic markup becomes more prevalent and the desirability of particular use cases becomes more clearly defined.

5. Where it is meaningful to do so, someone will convert available data to a standards-based semantic representation, probably for free.

Many current applications, whilst not conforming to semantic standards, do have data structures that can be described in RDF, FOAF, OWL etc. They also have feeds and apis that can be freely queried retuning XML, RSS, ATOM or JSON data that can be transformed. In an open environment, given sufficient interest, developers will write scripts or plugins that conduct these transformations so that the data can be added to the Graph.

6. The Graph will become ubiquitous.

As hand-held browsers, such as Nokia 810, iPhone and Android-based phones grow in popularity and capability, the online representation of our Graphs will become richer and more meaningful. An increase in the use of location and mapping services will stimulate not only improved geotagging by businesses, but also increased competition in price and service offerings as neighbouring competitive deals can be found.

With cameras and blogging software embedded in the device, product comments and reviews become far easier to transmit to our networks, making it easier to share images of that lovely, lovely cashmere.

Which of these steps will contribute significantly is a matter of guesswork and judgement. Only two things are certain: People will communicate, though the tools they use will change, and our Graphs will grow, however they’re expressed.