In the following sections, we show how the content of these captions can be derived automatically based on the interaction with complex information spaces.
The information space consists of structured models and related textual information. Structured models refers to geometric models that consist of distinct objects together with some semantic information, at least the names of objects and their affiliation to categories and subcategories. Furthermore, cross-references between objects (e.g. muscles and bones which are located near to each other) are represented.
The generation of figure captions that remain consistent with the image requires that all changes to a visualization are represented explicitly in data structures. As we pointed out in Section 4, figure captions also depend on textual components, e.g. on the availability of labels. Therefore, the visualization component consists of a text and graphics display (cf. the architecture in Figure 4).
Figure 4: Architecture of a visual interface with dynamic figure captions.
In order to describe the modifications that have been applied to the visualization, it is necessary to represent not only the cumulative state of an image and the corresponding textual elements, but also the interaction history. For this purpose, an agent is required which is informed whenever the visualization changes. This agent manages the context of the interface and is therefore called the context expert. The context expert communicates with the graphics display and the text display.
If figure captions are manipulated to control the visualization, a message is sent to the Context Expert (representing the requested changes) and the visualization. The context expert provides a full representation of the state of the visualization. This includes displayed labels and explanations, the viewing direction, colors, scaling factors of the displayed models and their objects together with the sequence of events, resulting in that visualization.
The interactive figure caption module analyzes the changes to the visualization and initiates the text generation based on the user's specification.
For interactive 3D illustrations, the most important aspect to comment concerns the visibility of objects after changing the viewing position. The 3D model has to be analyzed concerning what objects are hidden and to what extent as well as which objects are now visible. To describe the visibility of objects, a visibility analysis is employed.