Dec 05, 2012 (03:12 AM EST)
Semantic Databases: Destiny Or Distraction?
Read the Original Article at InformationWeek
Download the entire December 2012 InformationWeek special issue on semantic databases, distributed in an all-digital format as part of our Green Initiative
Don't buy into the idea that semantic database technologies are just for consumer-facing services such as BBC Online or the semantic Web initiatives embraced by the likes of Best Buy and Cisco. In much the same way that consumerization drives innovation in end user computing, semantic database technologies deliver benefits that businesses of all stripes should be exploiting.
At a high level, semantic databases offer five main benefits: They work with your existing relational databases. They align with Web technologies. Their underlying technology speeds integration of multiple databases. They're based on data structures that are flexible by design. And thus they can help enterprises tackle big data challenges.
Their biggest selling point is that data relationships are continually evolving. In the 1980s, we associated people with home and work phone numbers and a mailing address. By the '90s, we also had to represent cellphones and email addresses. And now we associate people with a broad spectrum of websites and social networking identifiers.
Connecting these identifiers isn't the end of the task. Our systems must understand the correct context for using this information. In other words, there needs to be machine-readable meaning associated with data.
Not coincidentally, the mantra for semantic technology is that information must be machine readable. The data definition must provide meaning and context so that a computer can understand the information rather than require a human to interpret it. This goes beyond supplying a structure for data, such as an XML schema. It requires that concepts within the data be given an ontological basis -- definitions that humans and machines can interpret.
The semantic technology standard for describing the meaning of data is the Web Ontology Language, or OWL (yes, the "O" and "W" are reversed on purpose). OWL defines a rich set of data relationship descriptors used to create a set of definitions for business terms, data sets and attributes. This resulting set is called an ontology.
Typically, a business will create and use multiple ontologies. At a corporate level, they define operations, departments, metrics and other core business concepts. At a system level, ontologies describe the types of data being managed and the specific attributes of that data, as well as the structure and relationships inherent in the attributes. Although separate, these ontologies are related, so that system-level details relate to corporate-level concepts.
Download the entire December 2012 InformationWeek special issue on semantic databases