Headup

For some time I’ve been meaning to write a post that would outline OUR view of the evolution of the Semantic Web.

What is the Semantic Web (aka “Web 3.0″)?

There are quite a few definitions out there for the “Semantic Web”, sometimes referred to also as “web 3.0″ – a phrase originally coined in 2006 by John Markoff of the New York Times. Two definitions worth noting canbe found on Wikipedia & the W3C.

Many of the definitions out there are excellent and inspiring and familiarizing ourselves with them was a necessary step for us in order to come up with a definition of our own, a definition that was formulated by our founder and CTO, Tal Muskal:

“The Semantic Web uses formats understandable by computers to describe the relationships that exist between things, concepts, people and other entities.“

“Formats understandable by computers” – the importance of Metadata

One of the greatest capabilities of our human brain is its ability to infer context:

• We see a person smiling and infer they are happy.
• We see a description on a web page and infer that it relates to the picture next to it.

Computers are unable to infer context. Their prodigious computing and logic capabilities are stumped by the smallest requirement for intuition. In this at least we are still superior to our digital progeny. However it is the ability to infer context that is the prime requisite for the emergence of the “Semantic Web”. Without the ability to understand context it is impossible to describe the relationships between different entities.

In order to assist computers in understanding the relationships between different types of entities and data it is first necessary to enable them to understand and describe a single type of data. This is where “Metadata” comes into play. Metadata was conceived with exactly this task in mind and is essentially data describing data. Confused?  The following images serve as a good example of the relation between data and the metadata about it:

Data:

This picture is an example of "Data"

Metadata:

This is an example of "Metadata" - It's data about the picture above

Even from this simple example it is possible to infer how metadata makes it much easier for computers to “understand” data by providing more information about it.

How does Metadata assist computers in describing relationships between different entities?

The evolution of the Semantic Web is closely related to the progression and development of software that cross references metadata from different sources because this is the essential  foundation necessary for the development of the ability to infer context by computers. Comparing metadata from different data-sets allows computers to test for similarities between the metadata, similarities that infer a contextual link between the original data-sets.

If we return to the photo above the metadata about the location where the photo was taken can be used by a computer in order to cross reference the photo with data supplied by a location-based social network like brightkite.com in order to provide members of the network with photos of their surroundings. This is a classic application of “Semantic Web” – usage of  formats understandable by computers in order to describe the relationship between a concept, in this example a “place” called “Haifa”, and people – in this case the members of the brightkite site.