Instance-Based Data Management -- Is it Useful for Biodiversity Informatics Applications?

Traditional data models explicitly or implicitly assume that data are organized according to a single, "correct" classification scheme. However, there is increasing recognition that biological and other phenomena can be classified in multiple ways to accommodate varying perspectives. In this context, we review an approach to instance-based data modeling that might be useful for managing biodiversity-related data. We examine how data can be represented to support multiple classifications, and highlight a novel SQL-like query language to support information retrieval in instance-based databases. We conclude by speculating briefly on the potential use of this model to support the data requirements of biodiversity applications.