Chen-Chieh Feng and Yu Liang, 2010 Association of American Geographers' Annual Meeting(AAG 2010), Washington DC, 14-18 April, 2010.

Abstract: The paper investigates how data generated by various environmental sensors can be efficiently integrated for interdisciplinary modeling efforts. Specifically, it aims to examine an operational, ontology-based system that supports semantic queries, data retrievals, and provenance recordings for environmental modeling using data from disparate sources and domains. The research advances our understanding of how existing ontologies and ontology alignment methodologies can be adapted for the stated purpose. It also explores the system design that support workflows of environmental modeling tasks, which could involve exploring visually the spatial patterns of the sensed data.

To achieve the goal, we examine the domain knowledge from two groups working in the ocean monitoring and modeling fields, followed by interviewing domain experts from these two groups to obtain the concepts and the knowledge structure of the respective domains. We also examine the semantics of the data generated by these two groups and compile how these data are used among the scientists. We then investigate the degree the mapping between upper level ontologies and the domain concepts can be accomplished and the best strategy to align concepts from the two domains under investigation. Last, we examine how the implementation of such a system helps search and retrieve data with both high precise and recall.

Keywords: Ontology alignment, data integration, Ontology reasoning

Conference website: http://www.aag.org/cs/annualmeeting