120 likes | 192 Views
Radial Marking Menu Performance Improvement and User Type Detection. Tim Burke - tburke2@umbc.edu Prepared for CMSC601. Radial Marking Menus. Why?. Related Work - Design.
E N D
Radial Marking Menu Performance Improvement and User Type Detection • Tim Burke - tburke2@umbc.edu • Prepared for CMSC601
Related Work - Design • Gilles Bailly, Eric Lecolinet, and Laurence Nigay, “Flower Menus: A New Type of Marking Menu with Large Menu Breadth, Within groups and Efficient Expert Mode Memorization,” Proceedings of the working conference on Advanced visual interfaces, 2008. • Tobias Hesselmann, Stefan Floring, and Marwin Schmitt, “Stacked Half-Pie Menus: Navigating Nested Menus on Interactive Tabletops,” Proceedings of the ACM International Conference on Interactive Tabletops and Surfaces, 2009. • G. Julian Lepinski, Tovi Grossman, George Fitzmaurice, “The Design and Evaluation of Multitouch Marking Menus,” ACM SIGCHI Proceedings, 2010. • Krystian Samp and Stefan Decker, “Supporting menu design with radial layouts,” Proceedings of the International Conference on Advanced Visual Interfaces, 2010. • Feng Tian, Lishuang Xu, Hongan Wang, Xiaolong Zhang, Yuanyuan Liu, Vidya Setlur, and Guozhong Dai, “Tilt Menu: Using the 3D Orientation Information of Pen Devices to Extend the Selection Capability of Pen-based User Interfaces,” ACM SIGCHI Proceedings, 2008.
Related Work - Performance • Andy Cockburn, Carl Gutwin, and Saul Greenberg, “A Predictive Model of Menu Performance,” ACM SIGCHI Proceedings, 2007. • Amy Hurst, Scott E. Hudson, and Jennifer Mankoff, “Dynamic Detection of Novice vs. Skilled Use Without a Task Model,” ACM SIGCHI Proceedings, 2007.
Background - Predictive Model • “Morphing Menus” change over time in response to user • Yields a 12% to 25% reduction in selection time of frequently used commands even when factoring out user memorization
Background - Dynamic Detection • Dynamic user type detection through trained C4.5 decision tree statistical classifier • Novice and expert users differ in needs when interacting with software • Achieves 90%+ detection accuracy with proper training data
Experiment • Two experiments: • Menu adaptation through “morphing menu” concept from Cockburn experiment to radial marking menus • Build dynamic detection classifier from Hurst experiment by collecting training data and testing for accuracy with radial marking menus
Evaluation • Menu Adaptation Experiment • Statistically significant reduction in menu selection times as compared to baseline tests with static, unchanging menus • Dynamic Detection Experiment • Once trained, is the classifier able to detect the user as novice or expert with accuracy approaching that of the previous experiment (approaching 90% accuracy)
Conclusion • Radial marking menus poised to become more popular • Understanding ways to better leverage them to provide a more powerful user experience