Spatial Hypertext

an extended abstract
Catherine C. Marshall and Frank M. Shipman III

Hypertext, in the most general sense, allows content to appear in a variety of contexts. The immediate setting in which readers encounters a specific segment of material then changes from reading to reading or reader to reader. This powerful underlying concept is usually realized in both research and practical efforts within a node-link model: nodes are the holders of content, and links are the means by which the content is contextualized. Links are thus closely associated with navigation and mechanisms for traversal; they are a way to move from node to node.

Spatial hypertext has its origins in browser-based approaches in which the emerging hypertext network is portrayed graphically, in an overview; authors create new nodes and links from within this structural map. In browser-based hypertext, boxes generally symbolize nodes; lines represent the links among them. In a completely spatial view of hypertext, the lines -- links -- may be removed from the picture, and the nodes may move about freely against their spatial backdrop. Nodes may appear in different contexts through multiple references to the same underlying content.

We have found it is still possible for authors to express relationships among nodes by using spatial and visual cues like proximity, alignment, and graphical similarity [1]. For example, two adjacent or overlapping nodes might be related in a manner analogous to a simple binary link; nodes of the same color might belong to a common set; more complex visual structures might suggest a hypertextual composite.

We have developed a spatial hypertext system, VIKI, that supports the emergent qualities of hypertext structures [2]. Spatial hypertext systems like VIKI have unique expressive qualities that take advantage of the human perceptual system, spatial and geographic memory, and more generally, spatial intelligence. VIKI's visual and spatial structures facilitate exploration by both readers and writers. For readers, the system provides an opportunity to read in context, with awareness of the related, nearby nodes. For writers, the system supports the development of a visual language through informal interaction with a user interface similar to a graphics editor. A hierarchy of spaces and subspaces helps keep complexity tractable for both readers and writers; users traverse through levels of spaces that serve to localize context to a particular set of nodes.

Instead of requiring authors to construct explicit structures, the system incorporates structure finding algorithms that analyze the spatial layout and the visually salient properties of the information objects. Because structure may be left implicit to be recognized on demand, VIKI uses ready manipulability to give users a great deal of expressive flexibility, thereby addressing common problems associated with authoring hypertexts, such as the need to articulate tacit knowledge or commit to structure prematurely. In short, hypertext is created easily, perceived by readers and writers, and, if requested, recognized by the system and used as the basis for further interaction.

Figure 1 shows an in-progress VIKI information space. There are a number of text nodes, called objects, visible within subspaces, called collections. Many of the objects in this space, such as the one called "SigLink" refer to World-Wide Web pages, and can be viewed using a standard viewer; others are content that VIKI stores.

[1] Marshall, C.C., Shipman, F. M. III. "Searching for the Missing Link: Discovering Implicit Structure in Spatial Hypertext." In Proceedings of Hypertext '93, (Seattle, Washington, November 14-18), 1993, pp. 217-230.

[2] Marshall, C.C.; Shipman, F.M.; Coombs, J.H. VIKI: Spatial Hypertext Supporting Emergent Structure. In Proceedings of the ACM European Conference on Hypermedia Technologies (Edinburgh, Scotland, Sept. 18-23), 1994, pp. 13-23.