An SDD subset example for Data Interchange
P. Bryan Heidorn, Graduate School of Library and Information Science, University of Illinois at Urbana-ChampaignA variant of the TDWG Structure of Descriptive Data (SDD) proposed representation for descriptive data was integrated into two applications to explore the strengths and weaknesses of the approach. While the computer applications can insolate end users, such as taxonomists, from complexity, the details of the SDD must be addressed by the computer program developers and must capture the information needs of the taxonomists. This presentation addresses this issue of implementation complexity since this complexity can become a barrier to wide spread adoption of a standard. The first of the two applications that were developed is OpenKey. OpenKey allows taxonomists creating species descriptions for interactive keys to coordinate their character group, character, state vocabularies and images over the World Wide Web. The second application, the Biological Information Browsing Environment (BIBE), is a descendent of a system previously presented at this conference. BIBE is a combination full text search tool and state driven interactive key. BIBE was modified to index SDD style descriptions and generate interface characteristics from this structure. OpenKey maintains the descriptive data in a series of relational database tables for internal OpenKey use but uses a variant of SDD proposed standard to export data to BIBE and hopefully, other applications. SDD is not an accepted standard but this work is indicative of the representational format and operations that will be enabled by the standard. It also demonstrates some of the difficulties that will be encountered with its implementation.
The main difficulties encountered were complexity, and difficulty distinguishing vocabulary from format. The main problems with complexity arise from two sources, the underlying domain and the XML tools and syntax. There is little that can be done to avoid the domain complexity completely but these complexities will make implementation of any standard difficult. As examples, we will look at representational models for real number ranges and character images and species images. The verbosity and complexity of the XML standard and the rapid evolution of representational tools being built on top of the basic XML framework give application developers powerful implementation options but also are difficult for application developers to master.