Non-spatial and Geospatial Semantic Query of Health Information
With the growing amount of health information and continuous outbreaks of diseases, the retrieval of health information is given more concern. Machine understanding of spatial information can improve the interpretation of health data semantics. Most of the current research focused on the non-spatial semantics of health data, using ontologies and rules. Utilizing the spatial component of health data can assist in the understanding of health phenomena. This research proposes a semantic health information query architecture that allows the incorporation of both non-spatial semantics and geospatial semantics in health information integration and retrieval.
Ontologies and rules in this architecture are used to support several functionalities, such as matching the data with the same semantics from various sources and reasoning in the spatial dimension through spatial relations and spatial operations.
The geospatial component in the health data is incorporated in this study, and a geospatial-enabled approach has been proposed for semantic health information retrieval. The architecture we proposed applies ontologies, facts, and rules in health information reasoning and deduction from both geospatial and non-spatial dimensions. Ontologies and rules have been explored for the basic representation of health data from various sources in the Semantic Web.
We enable the representation of spatial component in the Rule Markup Language (RuleML), and a RuleML engine OO jDREW has been customized to support geospatial semantic query of health information.
Spatial relation and operation operators are also enabled in the OO jDREW engine for spatial reasoning and knowledge discovery. This ontology and rule based health information integration and retrieval architecture provides initial exploration on how to utilize both non-spatial and geospatial semantics for health information retrieval and the case studies has demonstrated how the semantic query system works.
Our case study on respiratory diseases has demonstrated the use of ontologies and rules to support automatic health information retrieval. Our future work will be on the enrichment of human knowledge as ontologies and rules for health data reasoning and deduction to make semantic query systems ready for real health applications.
The development of Semantic Web will further promote the representation of human knowledge as ontologies and rules for effective query of health information.
Figure 1. Architecture for semantic health information query.