Applying Ontologies and Semantic Web technologies to Environmental Sciences and Engineering

The complexity and diversity of knowledge and terminology within environmental sciences and engineering is one of the key obstacles for successful interdisciplinary studies. Relevant data is difficult to locate and retrieve primarily due to varying formats, schemas and semantics. For example, for a typical modeling assignment a researcher needs to acquire knowledge of individual computational models, search, gather and analyze raw data, ensure the high quality of data, transform the data into formats compatible to the computation models that he or she is to use and then finally perform the modeling. This process takes several days to months.

To address these problems, we propose to use ontologies and emerging Semantic Web technologies. Ontologies provide shared domain models that are understandable to both humans as well as machines. We used the Web Ontology Language (OWL) to define ontologies with the objective of improving data sharing and integration. These ontologies define several domain concepts and describe a variety of domain models being used within environmental sciences and engineering. Metadata ontology is developed to define every facet of environmental datasets. Its aim is to provide a conceptual schema for the dataset using the available domain ontologies. The overall goal is to achieve content based retrieval of datasets and integration of heterogeneous data. We demonstrate a few applications which use the developed ontologies to solve common environmental problems. Our results suggest that ontologies and Semantic Web technologies like RDF and OWL may provide the much needed semantics within these diverse domains of environmental sciences and engineering, and hence may serve as the building blocks for innovative solutions to existing problems.