Semantic design for unstructured health data: a review

Abstract

The healthcare sector today is largely driven by data; as such the need for information exchange between the various stakeholders is on the rise. To date various software applications have been developed to facilitate this exchange. However, there are emerging challenges preventing direct exchanges (interoperability) among healthcare software applications current in use in most developing countries. A large proportion of healthcare data is still in unstructured formats that are either not captured, or is poorly captured in structured data. Exchanging this type of data between hospitals and different healthcare applications is currently problematic since a clear well-defined method is yet to be put in place. This work therefore seeks to understand the most plausible method (s) that can be adopted to not only improve the exchange and retrieval of unstructured health data but also extract meaningful health knowledge from the largely fragmented and distributed unstructured health data without affecting the free text.