Whilst computers have long been used to process the structure (or 'syntax') of data, traditionally engineered software could not process its meaning. This meaning was implicitly bound up in the system code that processed the data. However, ontologies now make this semantic information available in an explicit and declarative manner. 

Representing the meaning of the data being processed provides software engineering with two key benefits: 

  1. Loose Coupling - Coupling describes the level of common knowledge required between two systems for them to share information. For this interaction to work previously, systems had to be designed to understand each other's data format and syntax. The prior knowledge built into the systems resulted in a close binding, or 'tight coupling'. 

    'Loose coupling' however takes the approach of developing each system to explicitly state its own requirements; without making too many assumptions about the other. This results in an architecture that allows for the replacement or changing of components without having to change other components in the systems.          

    To join loosely coupled systems, ontologies can be used to establish a shared semantic framework. As this semantic data provides knowledge of a meaning that's consistent across the systems, each system needn't know much about the other in order to exchange information effectively. 

  2. Semantic Reasoning - Previously, the analysis of a problem domain would be executed in the code specific to that domain. But by using ontologies, the concepts within a domain and the relationship between these concepts, can be represented. Having made the semantic data explicitly available, 'inference engines' can conduct automated reasoning. As such, ontologies allow software engineers to describe a problem domain and its specifically relevant ontological model, in a format that inference engines can use to arrive at logical consequences.  

These technologies and approaches are collectively known as 'knowledge representation', forming the third tier in the hierarchy between:  

  1. Data (at its most basic level)
  2. Information (adding context to the data)
  3. Knowledge (understanding how to use the information)
  4. Wisdom (understanding when to use the knowledge) 

Ontology’s unique approach will help CSPs by providing them seamless service modeling and service management capability in an agile, cost-effective manner.
Arindam Banerjee
Director, Enabling Technologies Service Provider Software Solutions
Yankee Group, 2009