Internet-based Biodiversity Database Workshop April 24 - 27, 2000

Data Management Issues: What's left to be said?

Doug Johnston

Director of the Geographic Modeling Systems Laboratory, University of Illinois Urbana-Champaign

During the course of the workshop, numerous presentations were made that indicate that the issue of creating a web-based Biodiversity Information System is not a technical one. Many efforts are underway or have completed that span a range of technical solutions from centralized to distributed systems, and from text-based to GIS-based. The prevailing sentiment expressed was that the principle barriers to success are institutional.

It is worth consideration that many technical innovations are created to circumvent institutional barriers (e.g. translators to overcome lack of standards) and the introduction of the web itself (through NCSA Mosaic) was a response to the difficulty of accessing internet services due to the many protocols in use.

With that in mind, this presentation offers a review of the current state of affairs from a technical perspective through review of several projects, but in particular the Multistate Aquatic Resources Information System (MARIS) and the prospects for the second generation internet currently under development. Advantages of the web-based databases developed in the first generation include: a common point of entry, variations in systems/locations are transparent to the user, access to the system is enforceable, etc. Major disadvantages include a tight coupling between the interface and the underlying data structures, poor efficiency and scalability, loose coupling with metadata, etc.

Second generation internet technologies will address many of these issues through projects such as GLOBUS, XML, and numerous other technological innovations involving distributed systems, effective interfaces, etc. The principle issue of institutional barriers remains, however. Until vested parties agree to common definitions, designs, and objectives, technology will more often than not be implemented to address varied human behaviors.