Yale Licenses Software Technology for Geo-referencing Specimen Records

The BioGeomancer project team at Yale University’s Peabody Museum of Natural History has been granted software licensing by Inxight Software, Inc. to develop a universal system for geo–referencing of the diverse specimen records that exist in natural history collections.

BioGeomancer is a research consortium of seven academic institutions that is coordinated by the University of California at Berkeley. The consortium will feature an online workbench and web services to provide geo–referencing for collectors, curators and users of natural history specimens.

Reed Beaman, associate director of the Informatics Program at the Peabody Museum, will lead Yale’s team in developing software tools to allow natural language processing of archival data records that were collected in many different formats.

Over the past 250 years, biologists have gone into the field to collect specimens and associated environmental information documenting the range of life. The results of these explorations are an irreplaceable archive of Earth’s biological diversity that plays a fundamental role in generating new knowledge and guiding conservation decisions. Yet, roughly one billion specimen records, and even more species observation records, remain practically unusable in their current form.

“The greatest challenge to the effective use of this continuously growing archive is geo–referencing or converting the text descriptions of places where data and specimens were collected (locality descriptions such as ‘North Beach, Point Reyes, Marin County, California’) into a formal geospatial coordinate systems that can be mapped,” said Beaman. “The number of unreferenced records has been overwhelming, and severely limits the effective analysis of past and current species distributions.”

The system Beaman’s team is developing, using Inxight’s SmartDiscovery™ Fact Extraction module, will have the capacity to process thousands of sample entries in a few minutes, expanding the information that can mesh with data from other disciplines (climatology, geology, human impacts, etc.) for biodiversity conservation.

The Fact Extraction software can identify and extract items, events, activities and relationships from text data sources like specimen labels. Consortium users will be able to obtain relationships and facts of interest from search results; identifying relationships between subjects that are not obvious, and identifying trends over time or by location.

The project is funded by the Gordon and Betty Moore Foundation, established in September 2000 with an emphasis on outcome–based grant making. The foundation concentrates on initiatives tackling complex challenges in the areas of environmental conservation, science, higher education and the San Francisco Bay Area. For further information on Inxight Software contact Kathy Bentaieb kbentaieb@inxight.com.