With Gartner and Forrester both identifying semantic technology as a key trend to watch in 2013, the first question many folks are asking is, "What is semantic technology?" My goal here is to answer that question not in a rigorous, academic or historical way, but in the way I answer my family and friends over dinner.
At Cambridge Semantics we have a product (Anzo) built on Semantic Web standards, so I usually start there. Anzo allows our customers to combine information from widely different sources and formats into to an environment where it can be viewed and analyzed. Okay, you say, no magic there: many software products do similar things. So, we need to dig a little deeper to understand the value:
- The information sources can be really varied: from Google News to Twitter to enterprise software to Big Data systems
- When we say "combine", we don't just mean in the same place, we mean combined into the same conceptual model, regardless of the source—linked together based on common concepts
- The information is represented in human understandable form; the concepts and relationships are ones we all use every day (subject, predicate, object - a car has an engine)
- Through the model, all of the information is available in an intuitive way to be searched, analyzed and visualized, and so we can ask questions we did not plan for up front
- When we want to add a new source of information we just add it and it becomes part of the existing model, without the traditional weeks of design or coding.
The semantic data model makes all of this possible. It is constructed of simple, human-understandable "sentences"—subject, predicate, object. By linking these sentences together, we can create a conceptual model. But, this is not just the conceptual model, it is also the way the data is stored.
So, at its core, semantic technology is a very simple but very powerful concept. At Enterprise Data World, one attendee called our demonstration "magic"... While it is most definitely not magic, the power and flexibility of conceptually linking data from disparate sources must be seen to be believed. This is truly breakthrough technology with significant implications for large enterprises.
