paper reading #6:Improving Command Selection with CommandMaps

Intro:
title:Improving Command Selection with CommandMaps
reference information:Austin,Texas
author bios:Joey Scarr†, Andy Cockburn†, Carl Gutwin‡, Andrea Bunt. All three come from different Unviersities, Unverisity of Canterbury, University of Saskatchewan, and University of Manitoba. Their majors are all the same: Computer science.

Related Works:

1. Measuring the true cost of command selection: techniques and results
2. Benefits of merging command selection and direct manipulation
3. Toolglasses, marking menus, and hotkeys: a comparison of one and two-handed command selectiontechniques
4. Integrating laboratory and field study for improving selection: Development of a battery for predicting air traffic controller success.
5. Marker-assisted selection to improve drought resistance in common bean
6. mproving command and control speech recognition on mobile devices: using predictive user models for language modeling
7. Speech-based cursor control: understanding the effects of target size, cursor speed, and command selection
8. Design and analysis of delimiters for selection-action pen gesture phrases in scriboli
9. Improving the execution speed of compiled Prolog with modes, clause selection, and determinism
10. Using task context to improve programmer productivity

The related work talks about how to use command selections to improve programmer speed, or other speed of users. However, the topic of this paper I read is very novel according to what I have read. It talk about how to improve the command selection, and in other works that they try to combine command selection to improve performance. This is related, however, it's totally different idea. It's not common to see this topic in other papers. In one related work, it talks about benefits of merging command selection and direct  manipulation. This is very command in other related work, but this paper have totally different idea. 

Summary:
Current GUI application's commands are design in hierachal structures. This increase the time to find a specific command for expert users. And they found there is way to improve this performance by CommandMap. This technique has two main properties: they show all of an application's commands at once, and they do so in a spatially stable fashion, allowing users to build up spatial memory of frequently used commands.They have integrate this CommandMap within Microsoft Office Word. As the research goes on(details on evaluation section), they found out there is not much effiency for novice users; however, much faster result for expert users.

Evaluation:

Their overall hypothesis is that spatial memory can be the basis for command-selection interfaces. The first study consist 12 participants from local university, and ask to perform three tasks. The first one is to determine familiar commands which is to inspect the study system's mock up of the Word 2010 interface and to indicate which Ribbon-based commands they were familiar with. Second tasks is to specify locations with Ribbon hidden, and last task is to select commands using the Ribbon. The result founding is for many commands, people do have a spatial memory of the commands' locations in the GUI were known to within 100 pixels. And other founding is people know the tab location of most of their familiar commands. And there is another main experiement which consists 18 participants, and they were ask to use Commandmap and normal menu with similiar task I described above. The result showed speed increase for expert users of 34% over menus and 25% over Microsoft's Ribbon, and the other showing no significant performance differnce for novices. They used unbiased measuers and quantitative measures according to what I showed above. The methods they use is systemic because it did many studies over this topic.

Discussion:
I do think this work did contribute to many other areas. In the related works I read from, if this people use CommandMap and other works together, then the speed of computing would be much faster for novice and expert users. The contribution is very novel in this area, since other works didn't talk about it.I am really intereseted in this topic that I hope I can do something related to it in the future.