The SheafSystem libraries implement the sheaf data model, a non-relational data model designed for representation, manipulation, and storage of the large and complex data sets common in scientific and technical computing. These data sets represent the topological, geometric, and algebraic information required by mathematics-based scientific and technical computing applications.
These data sets are much more structured than the “unstructured” data that is the focus of the “noSQL” category of non-relational data management technologies and also more structured than the tables of the relational model. The SheafSystem uses advanced mathematics – posets, lattices, sheaves, and fiber bundles – to revolutionize data management for this highly structured region of the data complexity spectrum.
Mark Verschuren of Shell International E&P Inc. demonstrates the proof-of-concept Structure Builder application, implemented using the SheafSystem APIs.