Broad, Interdisciplinary Science In Tela: An Exposure and Child Health Ontology

Authors
McCusker, Jamie
Rashid, Sabbir
Liang, Jason
Chastain, Katherine
Pinheiro, Paulo
Stingone, Jeanette
McGuinness, Deborah L.
ORCID
No Thumbnail Available
Other Contributors
Issue Date
2017-06-25
Keywords
CHEAR (Child Health Exposure Analysis Repository)
Degree
Terms of Use
Full Citation
Abstract
Data curation for interdisciplinary collaborative science requires a new online web-based approach that integrates domain knowledge from multiple resources and enables in tela (in the web) interactive collaboration between data providers, domain specialists, and data analysts. e Children’s Health Exposure Analysis Resource (CHEAR) is a resource for child development and environmental exposure data. e CHEAR Data Center has developed an ontology that integrates study and exposure data in a way that is consistent across the program, and integrates with many best practice relevant vocabularies and repository schemas. is includes the World Wide Web Consortium’s recommended Provenance Ontology (PROV), Semanticscience Integrated Ontology (SIO), the Chemical Entities of Biological Interest (CheBI) ontology, the Uberon multi-species anatomy ontology, and the Units Ontology as the starting point for our domain modeling. We mapped terms where they overlapped and extended these ontologies with classes that were required to support modeling and integrating data from epidemiology and chemical exposure measurements that comprise the majority of the data recorded by the CHEAR data center. In response to this challenge, we used an on-demand approach to develop the ontology based on a set of representative pilot projects in CHEAR. A er initial development, we evaluated the ontology for completeness in representing an additional pilot study. An epidemiologist was able to produce a mapping of the project to the ontology with only minor corrections needed by an ontology expert. In the large dataset that was tested, one third of the classes needed to represent the dataset needed to be added to the ontology, all of them in areas where we expected to see more ontology expansion. Our overall approach is to drive towards completion of coverage while being relatively easy to use for domain experts. Ultimately we aim to have domain experts handle the majority of extensions and evolution with small interactions with ontology experts. In this paper, we report on our on-demand approach for web-based collaborative interdisciplinary ontology development and maintenance and also introduce the resulting extensible and interoperable exposure and child health ontology.
Description
Department
Publisher
Relationships
https://tw.rpi.edu/project/CHEAR
Access