Keep everything as simple as possible; but no simpler. (attributed to Albert Einstein)
S3Model is an agile and open framework for computable semantic interoperability in healthcare and other domains. This capability is required in order to advance the usage of clinical (and other) decision support services (DSS) and Big Data analysis across information coming from multiple points. These points of data creation are increasing every day as a result of more institutions bringing new digital capabilities to more functions, in addition to the plethora of new devices arriving in the personal healthcare marketplace.
S3Model is not a standard in the historical sense of the term. The standards development process is far from agile; information management is too dynamic to wait for the 20th century standards processes. In addition, managing information across domains is too broad for a top-down, one-size-fits-all solution for information interoperability. S3Model is based on widely used standards for Linked Data exchange and expression on the semantic web.
S3Model uses these tools in conjunction with the Unix/Linux model of iterate and release often. A new release does not invalidate the previous models or data.
In designing S3Model we took the view that data is produced by people that have a need to record that data. Often that data is useful to others. In domains like healthcare and bio-sciences this is certainly true. To transfer raw data is not enough for the receiver to determine if the data is useful for their needs. They need to know much more about the contexts (ontological, spatial, temporal) in which the data was captured.
If I have a leaky faucet in my home and need a washer to repair it. I go to the local hardware store to purchase a new one. The attendant there doesn’t know what I need because they do not know what type of faucet I have. However, I can look at the package information and determine based on thickness, inside and outside diameters and shape (flat, conical, round, etc.) if a particular washer fits my needs. In this sense, S3Model is the packaging for data. It allows the data consumer to determine if the data is useful to them as is or maybe it can even be adjusted because the consumer knows the context in which the data was captured. Of course there are certain base standards in the plumbing industry. Just like we have datatypes (string, integer, dateTime, etc. ) and XML constructs (element, complexType, etc.). But the overall concept is similar.
S3Model can be implemented as the information model for applications ranging from mobile apps to enterprise suites. S3Model is also useful to enhance semantic interoperability of existing systems and healthcare IT interfaces such as those based on HL7 messaging. We have developed tooling to enhance existing CSV data files.
S3Model specifically addresses the need to communicate syntactically sound data with the intended semantics. The larger, point to point issues of information exchange are already handled by other formal and de-facto standards such as SOAP, REST, etc. and S3Model is fully capable of utilizing systems built on those standards.
S3Model is the foundation for retro-fitting your current applications to inter-operate with your future applications using the semantic web tools and Linked Data, without requiring wholesale replacement.
The research and development of S3Model is based on more than a decade of interation and implementation. There have been two completed MSc. projects and a PhD. project executed in the development and validation of academic foundation known as the Multi-Level Healthcare Information Management (MLHIM) project. A number of peer-reviewed publications are available in the archives at