All organisations have disparate, disconnected data systems that they need to utilize. These data systems can have varying technologies, data structures and underlying semantics, all of which can make interoperation difficult.
There are several levels of interoperability:
- Platform Level: The platforms on which the information systems are deployed. For example, Windows or Linux. The platform version can often be significant also - the different potentially un-interoperable versions of Linux, being one such example.
- Software Level: This level defines the software and applications which the information systems are implemented with. For example a particular software vendor's content management system, or indeed version of it.
- Encoding Level: The representation or encoding level. XML has effectively taken over this level as being the de facto means of encoding information structures. API's are also relevant here, and in particular the growing use of protocols such as SOAP, UDDI, and use of other Web services technology.
- Structure Level: Once you know how to encode the information, its conceptual data types and structure must be defined. Frequently an existing schema or standard can be used in whole or part to achieve this. As per the encoding level, the use of XML Schemas is frequently applicable.
- Semantic Level: Defining the information so it can be programmatically manipulated is important - as in the above levels - but what is even more important is the meaning or semantics of the information. Unless this is defined or at least understood, any developed systems (IT or otherwise) will fail. The use of metadata mapping, data dictionaries and semantic definition languages such as RDF are appropriate here.
- Architecture Level: Any interoperability framework needs to have a macro-level architecture. This describes the principles and philosophy under which the overall information architecture and information systems operate.
- Human Level: Last but by no means least is the human or community level. Any system must operate within the context of a community of stakeholders. Skill levels, training and expectations must be considered, as must the wider issues of the team, department and enterprise.
|
|
Recent interoperability projects
|