Within Information visualization (Infovis) systems, there are two main components: representation (the graphics, mapping from data to display) and interaction (human-computer interaction). Interaction rarely is the main focus of research efforts, yet is essential for Infovis.
In Infovis, we view interaction techniques as the features that provide users the ability to manipulate and interpret representations. The study determined that the best way to classify techniques would be through user intent. The following seven categories emerged: Select (mark something as interesting), Explore (show something else), Reconfigure (show a different arrangement), Encode (show a different representation), Abstract/Elaborate (show more/less detail), Filter (show something conditionally), and Connect (show related items).
Select: Provide users the ability to mark a data item(s) of interest to keep track of it. Select interaction is coupled with other interaction techniques to enrich user exploration and discovery.
Explore: enables users to examine a different subset of data. Due to limitations (large scale data, view/screen limitations, perceptual/cognitive human processing limitations) only some data can be viewed at one time. Examining other data helps gain insight.
Reconfigure: provides users with different perspectives on the data by changing the arrangement of the display. Sorting, rearranging, switching between 2 and 3D, moving nodes, and reducing obstructions are all examples of reconfiguring.
Encode: enables users to alter the visual representation including appearance (color, shape, size) of each data element. Changing chart types, adding or modifying color, and altering size, orientation, font, and shape are all used (sometimes together) as encoding.
Abstract/Elaborate: provides users the ability to adjust the level of abstraction, from an overview down to details of individual data and many levels in-between. Zooming, (not) displaying sub-trees, and hover over detailed information are all types of abstract/elaborate.
Filter: enables users to change the set of data items based on specific conditions, specifying a range or condition so only data meeting those criteria is presented. Query controls (filtering by attribute) are an example.
Connect: used to highlight associations/relationships between data items and show hidden data items relevant to a specified item. Connect interactions are often used with multiple view representations (showing corresponding data), highlighting connected nodes, or expanding a node so related nodes are added.
This paper draws attention to the importance of interaction in Infovis research, and expands upon its complexity. It also provides a user intent-based categorization to deal with interaction techniques in Infovis to help provide a framework for a true science of interaction.
Link to PDF