Towards a Content-Based Material Science Discovery Network

Many publicly available databases exist for managing materials scientific data. These databases contain information from a wide variety of work, and the information is typically encoded in some proprietary format aimed at highlighting the goals of their particular backgrounds and purposes. In order to accelerate the rate at which new materials are discovered, these databases must be federated to provide materials scientists with the means to efficiently access large quantities of highly relevant data. This position paper advocates the design of a content-based material science discovery network that can allow for more intelligent reasoning over the databases than current implementations can afford. We will discuss the gains of using a hierarchical ontology for describing metadata that captures the various layers of the materials science domain. We will then discuss our approach in a content-based networking context.