QuickSelect: History-Based Selection Expansion
ABSTRACTWhen editing a graphical document, it is common to apply a change to multiple items at once, and a variety of tools exist for selecting sets of items. However, directly selecting large sets can sometimes be cumbersome and repetitive. We propose a method for helping users reuse complex selections by expanding the set of currently selected items. We analyze a document's operation history to determine which items have been frequently edited together. When the user requests it, items that have been previously edited with the current selection can be added to it. The new selection can then be manipulated like any other selection. This approach does not require a semantic model of the document or relations between items. Rather, each expansion is based on what the user has done so far to create the document. We demonstrate this approach in the context of vector graphics editing. Results from a pilot study were encouraging. Reusing selections with pre-existing histories, users were more efficient at editing tasks with our QuickSelect tool. Subjective preferences from a usability study in a free drawing context indicate that selection expansion is easy for users to learn and to apply. FILESSara L. Su, Sylvain Paris, and Frédo Durand. QuickSelect: History-Based Selection Expansion. In Proceedings of the 35th Graphics Interface Conference, pp. 215-221, Kelowna, BC, Canada, May 2009.
Paper: PDF @inproceedings{Su:09:QuickSelect, author = "Sara L. Su and Sylvain Paris and Fr\'edo Durand", title = "QuickSelect: History-Based Selection Expansion", booktitle = "Proceedings of the 35th Graphics Interface Conference", pages = "215-221", year = "2009", } ACKNOWLEDGEMENTSThe authors thank the GI reviewers, the study participants, the Open Clip Art Library artists, Craig Scull for feedback on experimental design, and the members of the MIT Graphics Group for proofreading. Thanks to the developers of Inkscape for providing an excellent graphics editor upon which to build QuickSelect. Frédo Durand acknowledges a Microsoft Research New Faculty Fellowship, a Sloan Fellowship, and a generous gift from Adobe. This work was partially funded by the Singapore-MIT GAMBIT Game Lab. |