People on different sides of environmental issues rarely speak the same language, but Penn State researchers say information technology can bridge that linguistic gap.
Using computational models to explain people's understanding of the environment will facilitate the exchange of scientific information, said Fred Fonseca, assistant professor of Information Sciences and Technology (IST). And highlighting similarities between stakeholders' viewpoints could result in the development of public policies with more consensus.
"If we can extract different groups' models or conceptualizations, I think we would find some similarities, and we might become more aware of the ecological consequences of what we do," Fonseca said. "It might help us articulate more concepts, too, which also could help us make fewer mistakes when dealing with environmental issues."
Fonseca and co-author James Martin, associate professor of psychology, explained their approach in "Space and Time in Eco-Ontologies," a paper delivered at the American Association for Artificial Intelligence's Spring Symposium at Stanford University held March 24-26. The paper was given in the "Foundations and Applications of Spatio-Temporal Reasoning" symposium.
Fonseca notes how people expect the scientific community to be unanimous in its understanding over Rain Forest issues, for example. But Brazilian scientists' view of that sensitive environmental system may differ from American scientists' view on aspects that would seem undebatable such as the annual amount of deforestation. Similar differences exist among other stakeholders such as the Brazilian government, other South American governments and non-government organizations. Fonseca and Martin argue that computational modeling may make different groups' theories and data more understandable to other stakeholders.
"The key is to build a tool that describes the information and optimizes the process of sharing information," Fonseca said.
But before this kind of information sharing can happen among computerized information systems, people's concepts must be explicitly formalized. Ontologies are the explicit formalization of people's mental models.
As part of a National Science Foundation grant, Fonseca will apply ontology integration to understanding how the environment affects the spread of West Nile Virus in Pennsylvania. To build his computational models, Fonseca plans to interview state health department officials and members of the Penn State science community including meteorologists and ecologists.
"Maybe scientists understand the disease and the environment differently than those in the health department," Fonseca said. "By mapping people's ideas, we build a tree of knowledge, so that there are fewer problems between the information and the conceptual models."
*mah*
