Paper Reading #7: Performance Optimizations of Virtual Keyboards for Stroke-Based Text Entry on a Touch-Based Tabletop

Reference Information:Performance Optimizations of Virtual Keyboards for Stroke-Based Text Entry on a Touch-Based Tabletop by Jochen Rick.  Published in the UIST '10 Proceedings of the 23nd annual ACM symposium on User interface software and technology.
Author Bios:
Jochen Rick is currently a faculty member in the Department of Educational Technology at Saarland University.  He holds a PhD in Computer Science from Georgia Tech and spent 3 years as a research fellow at the Open University working on the ShareIT project.

Summary
Hypothesis
Keyboard layout plays a large role in the effectiveness of shape-writing, or stroke-based text entry.
Methods
A group of participants was asked to stroke through a series of four points.   The author collected the data for the time taken, distance, and angle between each point.  He then applied this data to an algorithm to determine optimum placement for letter in a layout based on frequency and order of use.
Results
After applying his mathematical model to a number of existing layouts, he found that wide-set keyboard layouts like Dvorak and Qwerty performed very poorly as swipe-text entry.  As he pointed out, however, this is to be expected as they were designed with a very different usage in mind.  On the other hand, he presented two optimized layouts, Hexagon OSK and Square OSK, which were markedly faster than their already-existing counterparts.  

Contents
The author begins by stating that there is a need to develop better layouts for stroke based text entry.  He describes the history of keyboard layouts and goes into detail about the efficiency and reasoning behind some of the more popular ones.  He then conducted a study on how users make use of touch-screen technology when entering text and applied his findings to a mathematical formula to determine an optimum layout for letters in a stroke-based entry system.  The two keyboards he came up with show an improvement in efficiency over all currently existing layouts.
Discussion

I found this article extremely interesting and I am very convinced by his findings.  If I had one disappointment, it is that he did not actually create and test his new key layouts on users.  From what I understand, his calculations of efficiency and effectiveness are based purely on simulations.  I think that this stroke-based text entry system may prove to be a better solution to touch screen text entry than traditional finger-tapping and I am interested to see its progress.