I honestly thought that the terms ‘Semantic Web’, and ‘Web 3.0’ had disappeared for good as concepts that belonged to an earlier phase of the development of the web. It wasn’t – and isn’t – that I don’t agree with the objectives of ‘semantics’; I am just a little unsure of the ideas that were associated with these terms at the time, regarding their context within the rest of the developing web. Semantics itself is a complicated area and there are three distinctive categories which outline the role of semantics in computer science:
- Operational semantics: the meaning of a construct is specified by the computation it induces when it is executed on a machine
- Denotational semantics: meanings are modelled by mathematical objects that represent the effect of executing the constructs
- Axiomatic semantics: specific properties of the effect of executing the constructs as expressed as assertions.
Why do I tell you this? Think Wolfram|Alpha and the recent upswing of interest in new ways to build search engines with more intelligence, or a semantic understanding built on having the computational power to deal with the calculations all of this entails.
Back to the Semantic Web, the general purpose definition of which is the embedding of additional semantic metadata within content in such a way as to allow a machine to construct a deeper meaning. This definition is something that usually brings up the topic of Web Ontology Language, or OWL. Some years ago, I got quite excited by OWL, and even blogged on it, but then the promised breakthrough didn’t seem to occur. Instead, the emphasis shifted from a web built around content and machines to Web 2.0, built around communications, interactions, and people who are usually pretty good at semantic understanding. I was therefore interested to see what a Web 3.0 Conference in New York in mid May might be about. The good news is that Web 3.0 now seems to have morphed into something more closely linked to the challenges of today.
Spot the change towards people from machines in the following description: The goal of Web 3.0 is to reorganize information so users can capture what things are and how they are related. This seemingly simple concept will have a profound effect at every level of information consumption, from the individual end user to the enterprise. Web 3.0 technologies make the organization of information radically more fluid and allow for new types of analysis based on things like text semantics, machine learning, and what we call serendipity — the stumbling upon insights based on just having better organized and connected information.
But why be interested in the topic at all? My answer is because it is increasingly an enterprise challenge to deal with the huge amounts of unstructured internal content now actively in use. Saying ‘deal with’ not only means the casual browsing and collection of content that individuals find useful for their own purposes, but increasingly, the focus on how this is being incorporated into and mixed with the authentic, internally generated data from the enterprise’s own systems. This represents a double whammy; on one side, the increasing business focus on collecting and importing content from external sources on events and markets, used to optimise their own positioning and sales; and on the other, as a response to this, the equally increasing attention that auditors are paying to the quality of data.
I have deliberately used the term ‘content’ to signify anything that does not have a trusted, authenticated, provenance, as opposed to ‘data’, meaning that which the enterprise has either generated from its own recognised sources and meets these criteria, or has established these criteria for the content in question and therefore accepts it into the enterprise database(s). Just consider the act of accepting content into an enterprise data base, it means categorisation and taxonomy rules have been used to position this data in accordance with the internal structure. That structure is in accordance with the recognised and understood semantics of the enterprise and its peoples, , but these semantics are very focused on certain clear topics that are of value to the enterprise and its activities. It’s not the challenge of creating an ability to handle semantic understanding across the entirety of the World Wide Web!
In fact, what is increasingly happening is that enterprises are creating for themselves a subset of the Web for their internal use, calling it wikis, blogs, social networks, using it to support their people in finding and using expertise and the related content in a relatively free flowing form. The argument is that the act of capture and classification actually restricts and constrains the value to one set of circumstances, as well as being expensive. At this point the value of some of the new tools available for semantics and some of the papers, case studies and discussions at the Web 3.0 event suddenly look interesting. Internal semantic management helps both sides; the people can find and follow trails in a more relevant contextual manner and those responsible for compliance and audit can also see the relationships in a manner more related to that of the internal structural data.
If you are looking for a good source of more information on the subject, try http://www.semanticweb.com/. In particular, to follow up on the way that the Semantic Web is now linking up with people, take a look at this article on applying semantic search to social networking. It’s early days once again to say the evolution is here, but the new approaches to the Semantic Web definitely make it increasingly relevant to the way we are using the web today in enterprises.