Introduction: Acknowledgements
From DSL
Support for the development GeNIe and SMILE has been provided in part by the Air Force Office of Scientific Research under grants F49620-97-1-0225 and F49620-00-1-0122, by the National Science Foundation under Faculty Early Career Development (CAREER) Program, grant IRI-9624629, by Hughes Raytheon Laboratories, Malibu, California, by ARPA's Computer Aided Education and Training Initiative under grant N66001-95-C-8367, and by the University of Pittsburgh Central Development Fund.
The past and present principal developers of GeNIe, SMILE, and SmileX (listed alphabetically) are:
- Steve Birnie
- Nancy Jackson
We would like to acknowledge contributions of the following individuals (listed alphabetically) to coding, documentation, graphics, web site, and testing of SMILE and GeNIe: Kimberly Batch, Avneet S. Chatha, Cristina Conati, Roger Flynn, Abigail Gertner, Charles E. Grindle, Christopher Hall, Christopher A. Geary, William Hogan, Susan E. Holden, Margaret (Peggie) Hopkins, Jun Hu, Kent Ma, Robert (Chas) Murray, Zhendong Niu, Shih-Chueh (Sejo) Pan, Bharti Rai, Michael S. Rissman, Luiz E. Sant'Anna, Jeromy A. Smith, Jiwu Tao, Kurt VanLehn, Martin van Velsen, Anders Weinstein, David Weitz, Zaijiang Yuan, Jie Xu, and many others.
Students in the courses Decision Analysis and Decision Support Systems at the University of Pittsburgh, Decision Support Systems for Public Managers at Carnegie Mellon University, and Decision Support and Expert Systems at the University of Alaska, Anchorage provided us with useful feedback and suggestions.
GeNIe and SMILE embed a number of good ideas that we have gratefully assimilated over time from other software, whether decision-theoretic or not. The great user interface of Analytica has been an inspiration and a role model for us. Analytica's user interface has been developed by Max Henrion and Brian Arnold at Carnegie Mellon University in late 1980s and early 1990s. Our treatment of submodels is essentially the same as in Analytica. Knowledge Industries' WinDX software has been a source of inspiration for our diagnostic functionality and interface. The ideas contained in WinDX, on information gathering modes, such as discriminating among groups of hypotheses and pursuing specific hypothesis, were developed over a span of time from the mid-1980s to the early 1990s by David Heckerman, Eric Horvitz, Mark Peot and Michael Shwe. The use of alternative abstractions of the differential diagnosis in value of information computations was pioneered in the Pathfinder project that was commercialized as Intellipath and then, refined later in the KI Bayesian network inference tool kit. Beyond functionality, the configuration of panes and bar charts for abnormalities, observations, and valuable tests in the KI software were valuable inspirations for the interface of GeNIe and SMILE.
We would like to thank the U.S. News and World Report for considerable data collection effort and generosity in making the collected retention related data available. This data has been used in tutorials for the learning component of GeNIe.
