Hunter Hale

This paper presents a novel method of generating a user interface via machine learning.  The basic requirements for the interface are entered into the interface generation algorithm along with any physical constraints associated with the device the interface will be presented on.  A trace of typical user activity is also provide to the algorithm.  From this the algorithm present the user with series of questions.  The results of these questions and the earlier requirements and constraints are feed into a machine learning algorithm which applies linear optimization to generate a semi-final interface to present to the user.  One of the more interesting aspects of this algorithm is that it continues to run after the interface has been generated and it looks at how the user interacts with the interface to generate improved versions that optimize the most common tasks.

I found the use case they presented to be very interesting, they use the dynamic interface generation to create interfaces that have been adapted to specific users motor skill impairments.  The algorithm works almost exactly the same as in normal inteface generation except that instead of testing various inteface configurations to see which ones the user prefers it is looking at which set of interfaces the user has the easiest time interacting with.  This allows it to build a custom set of interface tools that optimize on the level of motor control the user possess.  Both there statistical tests and there ancedotal evaluations presented very good results.

Bib:

Decision-Theoretic User Interface Generation
Krzysztof Z. Gajos and Daniel S.Weld and Jacob O.Wobbrock
Proceedings of the Twenty-Third AAAI Conference on Artificial Intelligence (2008)

Rodney Brooks

1986 IEEE

Brooks presented the initial research on the subsumption architecture here. Subsumption was originally designed to provide several benifits to the programers of mobile robots. It is supposed to allow the user to design modular reusable components, handle multiple sensor input, and deal with possibly inaccurate sensor results. The utility of subsumption was demonstraighted with the construction of both a physical agent capable of navigating a dynamic enviroment as well as a software simulation. The initial robot was only programed for navigation but additional tasks were planned to further show off the capabilities of subsumption. Subsumption was also shown to be very reusable and able to overcome noisy sensor data.

Link: A Robust Layered Control System for a Mobile Robot – PDF

Bibtex:

@article{basicSubsump,

Author = {Rodney Brooks},

Date-Modified = {2008-04-07 17:18:04 -0400},

Journal = {IEEE JOURNAL OF ROBOTICS AND AUTOMATION},

Number = {1},

Pages = {14–23},

Publisher = {IEEE},

Title = {A Robust Layered Control System for a Mobile Robot},

Year = {1986}}

by – Maja J Mataric

Summary This paper presents a method of developing a landmark based navigation method for robotic agents. The agents are programmed with a subsumption architecture capable of encoding information about things they have seen with there sensors to form a self improving navigation map. This development of a navigation map is preformed in real time using just the robots sensor systems. This was the first subsumption architecture to allow for spacial planning while remaining fully reactive and one of the first to implement the behavior controls.

*** I am not sure if this is the paper that presents the behavior based system fully. They seem to gloss over the implementation of the behavior system.

*** The agents generated by this system seem to have something of sense of place which might make them something interesting to look into for the DEACCON work.

Link:Integration of Representation Into Goal-Driven Behavior-Based Robots – PDF

Bibtex:

@article{bControl,

Author = {Maja J Mataric},

Date-Modified = {2008-04-07 17:18:04 -0400},

Journal = {IEEE Transactions on Robotics and Automation},

Number = {3},

Pages = {304–312},

Publisher = {IEEE},

Title = {Integration of Representation Into Goal-Driven Behavior-Based Robots},

Year = {1992}}

Summary:

This report is a well written and extremely thought out examination of the ethics and responsibilities of researchers who want to do work with human subjects and what standards they should be held to. The basic principles can primarily be condensed into 3 concepts: First being that there needs to be respect for the human test subjects. They need to be clear and true volunteers who have as full an understanding of the research they will be participating in as possible. Next the beneficence of the work needs to be examined. The risks presented to the study participants need to be minimized down as far as possible and the benefits to the participant or society need to be maximized. Finally the there is the question of Justice. The application of the research should be spread as far as possible and not limited to small groups and the actual performance of human trials should not be restricted to any small groups unless so dictated by the research itself.  Personally I found the first two principles to be kinda obvious however the third principle was moderately thought provoking.

Link: Official Site   Want to convert to pdf before making local copy

Bibtex: Coming