Design and realization of spectral data management in the spectral knowledge base

The measured spectral data in the spectral knowledge base (spectrum library) are all high spectral data, which are composed of band values and corresponding reflectance values, data explanation and self-contained parameters. How to realize management of those magnanimous data stored in database effectively, including presenting them out logically, eliminating data redundancy and so on, becomes an important problem to be solved. For this purpose, following key points about data management are introduced and discussed in this paper: (1) design of data operation flows; (2) implementation of data deletion and reversion piecemeal or in batches within mechanism of recycle bin; (3) redundant data cleanup periodically; (4) generating of reports about quantity and availability of data in storage respectively; (5) presentation of spectrum profile with corresponding metadata in various ways. On the platform of Oracle 9i, Windows 2000 Server and IIS, this solution to spectral data management was implemented and proved to be rather effective, as vas instanced by spectral knowledge base.