|Home||Archived February 20, 2019||(i)|
Moving Toward a World Wide Web for Scientific DataWorking Sessions on Use Cases for Semantic-Web Development
Imagine a dozen U.S. Geological Survey (USGS) employees surrounded by projection screens, marker boards, and flip charts. They are drawing “activity diagrams” with hangman-like stick figures and arguing over the meaning of words.
Online shopping and social networking have made great strides in recent years, largely through use of “semantic web” methods that allow computers to automatically interpret information, rather than simply displaying it. The USGS Coastal and Marine Geology Program (CMGP) is learning to use these semantic-web methods to increase the value of online scientific data. Scientific use of the semantic web is especially appropriate because these methods automatically integrate data from diverse sources.
The USGS recognizes that scientific progress and sound management of the nation’s resources increasingly depend on integration of data from diverse projects, not only from within the USGS but also from other agencies and research organizations. The semantic-web approach will allow the integration of CMGP databases, without significant modification to individual datasets, and facilitate integration of these CMGP databases with information systems being produced by the National Science Foundation (NSF), the USGS Council on Data Integration, the Interagency Working Group on Ocean and Coastal Mapping (formed in response to the Ocean and Coastal Mapping Integration Act of 2009), the National Ocean Council, and state and regional organizations engaged in coastal and marine spatial planning (http://www.whitehouse.gov/administration/eop/oceans/cmsp). In a broader context, this CMGP effort is consistent with the government-wide commitment to the semantic web (http://www.data.gov/semantic).
In September 2011 and January 2012, USGS employees met with experts from Rensselaer Polytechnic Institute (RPI) and Woods Hole Oceanographic Institution (WHOI) to learn about semantic-web development and begin implementing the semantic-web-development methodology employed at RPI and WHOI. The foundation of both meetings was a lecture by RPI professor Peter Fox on the RPI methodology for semantic-web development. Central to this methodology is a “use case,” which identifies a user’s goal and describes the steps or actions between a user and a software system required to achieve this goal. Analysis of the information and activities necessary to meet each use-case goal leads to development of an information model. The concepts and the relations between them that are expressed in the information model allow a system to be designed to meet many similar goals. Analysis of several use cases serves to identify the full requirements of the system. The RPI process is iterative, with rapid software prototyping, expert reviews, and ongoing communication among subject-matter experts, modelers, and software engineers. This methodology is being used to develop innovative scientific information systems by such organizations as WHOI, the National Aeronautics and Space Administration (NASA), and the NSF. The NSF-funded Virtual Solar Terrestrial Observatory (VSTO) is one early success story for RPI methodology (http:/www.vsto.org/; see also http://tw.rpi.edu/web/project/VSTO).
After the September lecture, participants applied the use-case methodology to their own goals for USGS information systems. Discussion of these first use cases, and comments provided by Professor Fox, clarified the types of description and level of detail that are needed for a use case to serve as the basis for an information model and for implementation by software engineers.
During the final stage of the September meeting and most of the January meeting, participants were divided into teams to work on specific use cases identified as especially important for the CMGP. These use cases describe specific, hypothetical scenarios in which given users achieve their information-seeking goals: for instance, a Massachusetts coastal-zone analyst seeking and obtaining integrated CMGP seafloor data, with the ultimate goal of characterizing benthic habitats. A few carefully selected use cases of this type, which, taken together, represent a wide range of users, information needs, and information-seeking behavior, can be used to guide information-system development.
The workshop teams were formed to include a mix of skills and roles. In addition to subject-matter experts who know about CMGP data acquisition, analysis, and publication, each team included a knowledge modeler, a software engineer, a facilitator, and a scribe. Through an iterative process, each team discussed and clarified the elements of the use case: goal, actors, preconditions, triggering events, postconditions, normal and alternate flow of events, and resources required. The teams also drew diagrams to summarize the interactions between people and the information system, and they developed models to show the relations among essential information concepts. Team members agreed to work on refining two use cases and implementing prototype systems to test them.
The working meetings were held at the USGS Woods Hole Coastal and Marine Science Center in Woods Hole, Massachusetts, September 21-23, 2011, and the USGS St. Petersburg Coastal and Marine Science Center in St. Petersburg, Florida, January 18-20, 2012. Leaders were Peter Fox (RPI and WHOI), Andy Maffei (WHOI), and Joe Futrelle (WHOI). Participants were Emily Himmelstoss, Seth Ackerman, Fran Lightsom, VeeAnn Cross, Brian Andrews, Ellyn Montgomery, Guthrie Linck, and Brad Butman of the USGS Woods Hole Coastal and Marine Science Center; Rob Wertz, Shawn Dadisman, Jamie Cormier, Theresa Burress, Kristy Guy, Brendan Dwyer, Heather Schreppel, Bryan McCloskey, Greg Miller, Karen Morgan, and Jolene Gittens of the USGS St. Petersburg Coastal and Marine Science Center; Alan Allwardt, Carolyn Degnan, and Rex Sanders of the USGS Pacific Coastal and Marine Science Center; Janice Gordon of the USGS Bioinformatics Program; Dahlia Varanka of the USGS Center for Excellence in GIS; and Massimo DiStefano of RPI and WHOI. Peter Fox’s lectures were shared via WebEx with the CMGP’s information-management community and the USGS Community for Data Integration.
in this issue:
|Home||Archived February 20, 2019|