1. local global disambiguation of terms and concepts The BCO-DMO metadata database uses controlled vocabularies to record many of the important pieces of information that document the data sets in the BCO- DMO database. To improve semantic interoperability, terms from the local BCO- DMO controlled vocabularies are being mapped to controlled vocabulary terms adopted by other oceanographic data management organizations. For example, we have mapped the BCO-DMO Instrument vocabulary to the 'SeaDataNet Device Categories' vocabulary (http://vocab.ndg.nerc.ac.uk/list/L05/current) served from the British Oceanographic Data Centre (BODC) Natural Environment Research Council (NERC) Data Grid vocabulary server (http://www.bodc.ac.uk/products/web_services/vocab/) 1. The SeaDataNet device categories vocabulary and related sampling and sensor type vocabularies met our criteria 2 for a community standard vocabulary: availability (Web-accessible, dereferencable HTTP URLs) quality (completeness, clarity and precision, relevance) community adoption effective governance structure References: 1. Lowry, R and G. Williams (in press) "Putting Meaning into SeaDataNet", Mediterranean Marine Sciences. 2. Graybeal, J. 2009. "Choosing and Implementing Established Controlled Vocabularies." In The MMI Guides: Navigating the World of Marine Metadata. http://marinemetadata.org/guides/vocabs/cvchooseimplement This work has been funded by the National Science Foundation. We are especially grateful for the significant contributions made by Peter A. Fox (pfox@cs.rpi.edu) at the Tetherless World Constellation and Rensellear Polytechnic Institute and BCO-DMO programmer, Tobias Work. The ontology concept map illustrates the semantic relationships between classes represented in the BCO- DMO data system. The ontology is available from MMI ORR at http://mmisw.org/orr/ and RPI TWC SVN site at http://escience.rpi.edu/ontology/BCO-DMO/bcodmo/2/0/. Challenge: to develop semantically-enabled search capability for the BCO-DMO data system (e.g. faceted search) Strategy: follow the Semantic Web Methodology and Technological Development Process described by Benedict et al. (2007) to develop an ontology for the BCO-DMO data The Challenge Abstract ID: EGU2011-8578 EGU General Assembly 3-8 April 2011 Developing an Ontology for Ocean Biogeochemistry Data Biological and Chemical Oceanography Data Management Office Cynthia L. Chandler 1, Molly D. Allison 1, Robert C. Groman 1, Patrick West 2, Stephan Zednik 2, Andrew R. Maffei 1 1 ~ Woods Hole Oceanographic Institution, Woods Hole, MA and 2 ~ Tetherless World Constellation, Rensselaer Polytechnic Institute, Troy, NY A discipline as diverse as oceanography will benefit greatly when the research community meets the challenge of effectively incorporating Semantic Web technologies into the existing cyberinfrastructure. Semantically-enabled data delivery systems offer great promise for enabling new and better scientific research, but significant challenges must be met before the full potential can be realized. Evolving expectations for open access to research data combined with the complexity of global ecosystem science research themes present a significant challenge, and one that is best met through an informatics approach, wherein research scientists, information managers and computer scientists collaborate in small teams. The Biological and Chemical Oceanography Data Management Office (BCO- DMO) is funded by the US National Science Foundation Division of Ocean Sciences to work with ocean biogeochemistry researchers in the US to improve access to data resulting from their respective programs. In an effort to improve data access, BCO-DMO staff members are collaborating with researchers from the Tetherless World Constellation (TWC) at Rensselaer Polytechnic Institute (RPI) to develop an ontology that formally describes the concepts and relationships in the data managed by the BCO-DMO. The project required transforming a legacy system of human-readable, flat files of metadata to well-ordered controlled vocabularies, and finally to a fully developed ontology. To improve semantic interoperability, terms from the BCO-DMO controlled vocabularies are being mapped to controlled vocabulary terms adopted by SeaDataNet, the pan-European infrastructure for ocean and marine data management. Additionally, as part of their efforts to develop generic science ontologies, the team at TWC has facilitated the adoption of key concepts from the BCO-DMO ontology into ontologies developed for other science domains, and the adoption of concepts from other domains into the BCO-DMO ontology. From the beginning of the project, development of the ontology has been guided by a use case based approach. The use cases were derived from data access related requests received from members of the research community served by the BCO-DMO. The resultant ontology represents the information stored in the metadata database and satisfies the requirements of the use cases. The BCO-DMO metadata database currently contains information that powers several different user and machine-to-machine interfaces to the BCO-DMO data repositories. One goal of the ontology development project is to enable subsequent development of semantically- enabled components (e.g. faceted search) to enhance the power of those interfaces and improve data access through enhanced data discovery. Abstract Use Case Small Team, mixed skills Analysis Adopt Technology Approach Leverage Technology Infrastructure Rapid Prototype Open World: Evolve, Iterate, Redesign, Redeploy Use Tools Science/Expert Review & Iteration Develop model / ontology Semantic Web Methodology and Technology Development Process Acknowledgments photo by Molly D. Allison Ontology Concept Map The process of matching and mapping supports use of: 1) local vocabulary terms, familiar to the originating investigator; 2) intermediate, consistent terms managed by repository custodians (e.g. BCO-DMO); and 3) closest match terms (e.g. from SeaDataNet), shared by the larger community. Multi-level matching and mapping enables retention of important information while improving interoperability of data systems. Reference: James L. Benedict, Deborah L. McGuinness, Peter Fox. A Semantic Web-based Methodology for Building Conceptual Models of Scientific Information, American Geophysical Union, Fall Meeting (AGU 2007) (Eos Trans. AGU 88(52), Fall Meeting Supplement, Abstract IN53A-0950), 2007. http://bcodmo.org Vocabulary Term Matching and Mapping to SeaDataNet Instrument Mapping Examples original dataBCO-DMOSeaDataNet SIO-CTDCTD Sea-Bird 911 Sea-Bird SBE 911 CTD /term/L221/current/TOOL0035 bongo towbongo net plankton net /term/L051/current/22 XBT Expendable Bathythermograph Bathythermographs /term/L054/current/132 see mapping table above right