Preserved plant and animal specimens in natural history museums and herbariums are critical resources for biological research. Systematists, ecologists, biogeographers, and investigators from many other fields rely on these specimens and the data associated with them. Most specimens are accompanied by information concerning their geographic location, but most commonly these locality data are in the form of narrative descriptions, rather than standardized map coordinates such as latitude and longitude. Dr. Matthew McGranaghan, a geographer, proposes to develop new computer software that will use an expert system approach to interpreting these text-based locality descriptions into map coordinates. The software will call on electronic gazetteers and digitized maps that are currently available. The proposed software will be tested on label data from the plant collections in the Bishop Museum of Natural History of Hawaii. Biologists are keenly interested in understanding the geographic distributions of organisms, but it is prohibitively time-consuming to plot individual specimens on maps. The proposed software development effort has the potential to make hundreds of millions of museum specimens readily available to researchers working on problems of geographic distribution. This effort will have wide impact in the museum research community, in geography and regional sciences, and in the development of artificial intelligence applications in biology.
