Just semantics? Or is it true?

The semantic web is one of those things that has been much talked about, especially in propellor-head circles, for several years. It leaves most outsiders a bit cold. They actually don’t care about triples, RDF or SPARQL. They care about whether it will make a difference to their businesses.

Well, after a circuitous journey which took in two reports (here and here*), two related podcasts (here and here) and a series of instant messages with, my semantic everything guru, Dr Paul Miller, I decided that one of the reports was fairly grounded and quite useful. It’s the (asterisked) one by David Provost, an analyst whose mid-life MIT SM thesis was called ’Hurdles in the Business Case for Semantic Web’.

Since then, he’s sustained his interest in the subject and recently produced the report (free to download) called On The Cusp: A Global Review Of The Semantic Web Industry. It starts with a plain English run down of the semantic web world and finishes with a look at 17 semantic technology vendors and, in many cases, their clients. He started out by inviting 25 companies that he figured were doing real stuff, but eight either refused or didn’t respond to his invite.

The resulting document verges on the gushing but, if you ignore that, it contains much of interest. It is strictly business-focused and it tries to steer clear of all but the most essential jargon.

A lot of the software described has been around for donkey’s years but it is finding its vocation in the semantic web world. Natural Language Recognition, for example, is a core requirement for ’reading’ and interpreting documents and turning the findings into the metadata which is at the heart of the computer’s ability to understand content and context.

Databases with extensible schemas are another key element. This is necessary because the world is always changing, we can’t predefine everything in advance.

The vendors vary in what they offer, from the narrowness of USA database company Franz, to the breadth of German company ontoprise GmbH. I quite like the sound of thought-sharing company Primal Fusion – but that’s still in alpha so I couldn’t check it out. I did look at Thomson Reuters’ Calais however, but wasn’t totally gripped by the results. It reminded me of speech recognition: no matter how good, it takes a while to get used to you. I rarely add Ltd or Inc to company names, so when I pushed some of my work through for automatic creation of metadata, some key elements, like company names, were not spotted.

It is, as the author points out, early days but enough is going on to suggest that we’re going to see increasing use of semantic activity to improve performance for the benefit of organisations and their customers.

If this is unfamiliar territory to you, then the report will start you off, regardless of whether you’re a business person or a propellor-head.

I would also recommend keeping a close eye on what Dr Paul Miller is up to. He seems to know everyone and everything that’s going on in the semantic space, web or otherwise.