Capturing Description Data for Taxonomic Projects
4.2 System Concept
4.2.1 Introducing the System
The basic system concept under investigation is one in which a taxonomist specifies the descriptive specimen data to be collected, in terms of a structured data model and defined terminology. This generates a structured proforma, which is used to control the data entry of specimen descriptions in a consistent manner for a given taxonomic project.
The high level task of capturing specimen description data can be broken down into two main user tasks:
• Specifying what description data is to be collected for the taxonomic project. (see Use Case Scenario 2, Level 2: ‘create proforma’ in Appendix A).
• Entering the instance description data for each specimen in the taxonomic project. (see Use Case Scenario 2, Level 2: ‘scoring taxonomic description for a specimen’ in Appendix A).
As the process of the detailed taxonomic sort is an iterative process, the user also requires the ability to return to the proforma creation stage to add extra data requirements to the proforma as needed.
Altering the descriptive data specified in the proforma when data has only been collected on a proportion of the specimens incurs a potential weakness in the quality of data recorded, as for example data entry decisions could be made on different options for different specimens. However removing this possibility altogether is not possible within the stricture of taxonomic working practice. Taxonomists confirmed that they could not necessarily know all character concepts that would prove to be of interest during a project before the process of collecting descriptive character data had begun. As time pressures on taxonomists work was intense there was certainly no time to start the whole process of collecting description data again when character concepts to be recorded were added or extended. The approach thus must allow for at least a limited degree of iterative working between the two main user tasks.
Each main user task has its own system-generated interface upon which the user operates (see Proforma Builder UI and Data Entry UI in figure 4.1). Linking the two interfaces in an integrated system will support altering the proforma during a project. Figure 4.1 illustrates this two-interface system for addressing taxonomists’ needs as initially conceptualised. Two UIs are shown in the figure, one for building an electronic proforma and one for data entry based on that proforma. The system generates these interfaces as required to display the details of defined descriptive terms and constraints on how to combine those terms. Presentation and dialogs were based on known user tasks and an implicit mapping from the data model.
Descriptive Terms & Constraints Electronic Proforma SPECIMEN DESCRIPTION DATA Plant Specimens build proforma Proforma Builder UI description Data Entry UI DESCRIPTION DATABASE database Taxonomist Taxonomist system
Figure 4.1: System to capture specimen description data in taxonomy. Two user interfaces are represented: the Proforma Builder interface and the Data Entry
The decision was made to keep the system independent of the description database to make it platform independent. In addition to general adaptability this decision would support the possible future use of the system on more mobile devices to capture data in the field. This would also avoid ties to a server, which would be advantageous as networks may not be available in some locations where taxonomists work with specimens such as herbariums. It was decided that the Prometheus II database [Paterson 2004] would be suitable to be used for this purpose during development and it was assumed specimen data could be exported in that format. This database was being developed to hold taxonomic descriptive data in a suitably rigorous format.
4.2.2 Description Data
Description data must be applied and utilised by both users and system to describe specimens in a controlled fashion.
To address the needs of ensuring clarity and comparability of the collected description data, it was decided to use only descriptive terms with clear definitions and controlling relationships between them. Taxonomic description data consists of characters of interest, which as seen in chapter 2, can each be broken down into a physical plant structure, an aspect or property thereof and the actual state or value. Using this breakdown, an electronic proforma thus consists basically of the structure terms in which the user is interested, the aspects or properties of those structures they wish to comment upon and the possible values or states that an individual specimen could possess. A specimen description instance differs from a proforma primarily in that it contains the actual value/state that applies for that specimen from the domain of possible values/states.
The terms from which the proforma is built and the constraints on how these terms are used are integral to the system (see figure 4.1 ‘Descriptive Terms & Constraints’). The defined terms were organised as a glossary of defined descriptive structure, property and state terms to which the taxonomist could add (within certain limits) where necessary. The provided interface would control how these defined terms could be combined through a series of in-built rules. It was conceived that these rules would be loosely based on the Prometheus II data format rules [Paterson 2004]. The Prometheus II data format being developed in parallel with this project aimed to develop a database
model for taxonomic descriptive data. Unlike other available electronic description formats, it was intended to support comparability and was designed for use with a defined terminology. The details of the Prometheus II data model used during this stage of research can be found in Appendix C.
4.2.3 Presenting the interfaces
Figure 4.2 illustrates an example of the envisioned proforma builder UI. The panel on the left side is an overview of the current state of the proforma, representing the plant structures with the specified characters. On the right side is a definition explorer panel, where defined terms from the glossary are displayed, with filter, grouping and order buttons to control the view. A dialog pane for adding a character can be seen in the centre of the screen. The panels are co-ordinated with linking techniques.
Figure 4.2. Example of Proforma Builder interface storyboard.
Figure 4.3 illustrates an example of the data entry interface. The overview of the proforma is again seen on the left side of the screen, with a character presented for data entry on the right. Variations on interface presentation were investigated and are discussed at 4.5.
Figure 4.3 Example of data entry interface storyboard.