http://www.cra.com/bnet BNet.News
In This Issue: April 2005 
•   See how Scientists at Charles River used BNet.Builder to Create Models of Social Networks at the Conference on Intelligence Analysis, May 2-6
•   BNet.EngineKit Beta Test Program
•   Northrop Grumman Chief Technologist Uses BNet.Builder to Teach Himself about Bayesian Networks
•   Fun and Educational: Computerized Tongue Diagnosis Based on Bayesian Networks
•   User Tip: Easy Network Layout
•   Publicize Your Belief Network Success Story
See how Scientists at Charles River used BNet.Builder to Create Models of Social Networks at the Conference on Intelligence Analysis, May 2-6
Charles River Analytics Senior Scientist Jonathan Pfautz will present the results of work on modeling the effects of cultural and organizational factors on individual and group behavior at the International Conference on Intelligence Analysis on 2-6 May 2005. Among other tools, the work uses BNet.Builder and BNet.EngineKit to create and run these specialized social models.

Air Force Research Laboratory (AFRL) and the Natick Soldier System Center (NSC) sponsored this work because intelligence analysts now need tools to help predict the behavior of opposing forces using non-traditional, asymmetric warfare methods in culturally complex environments. The US Department of Defense is exploring increasingly sophisticated behavioral models and simulations for intelligence analysis. However, these models fail to effectively address some of the "soft" representational issues, including modeling the influence of culture on groups and individuals, how organizations can influence individual behavior and vice versa, the quantifiable differences between individuals, and the influence of situational factors on all of these. This already difficult problem of behavior prediction is exacerbated by the general lack of specific data about individuals or groups, and the uncertainty and unreliability of the available information that is inherent to asymmetric warfare.

See the Conference Agenda

Northrop Grumman Chief Technologist Uses BNet.Builder to Teach Himself about Bayesian Networks
Kevin Baxter, Chief Technologist for Systems Analysis and Engineering at Northrop Grumman Corporation, recently downloaded the trial version of BNet.Builder. His company is preparing to write a proposal for a project that will use Bayesian networks for an important client. Kevin was looking to find out as much as he could about building and using belief networks - quickly.

With the Charles River's introduction to belief networks, "About Bayesian Belief Networks", and a copy of BNet.Builder, Kevin taught himself what he needed to know. Within a week of downloading BNet.Builder, Kevin created an example network that predicts whether or not his children will stay home from school.

Download a copy of "About Bayesian Belief Networks"

User Tip: Easy Network Layout
Two buttons in BNet.Builder can help you create clean, attractive, organized networks: Magnetic grid and Hierarchical layout.

Magnetic grid To bring up a grid of dots, click on the Magnetic grid button. When Magnetic grid is on, nodes will snap to the grid when you create or move them, making lining up nodes easy and fast.

Hierarchical layout does all the work of lining parent and child nodes up so that child nodes are below the parent nodes upon which they depend. Clicking on this button will automatically adjust the layout of the entire network. If you don't like the results, just select "undo" from the edit menu (or use the control + Z keyboard shortcut).

BNet.EngineKit Beta Test Program
Buy your copy of BNet.EngineKit Revision 1.0 before the release, and be eligible to recieve a free copy of the full Beta version.

Find Out More

Fun and Educational: Computerized Tongue Diagnosis Based on Bayesian Networks
Researchers at the Biocomputing Research Center of
Harbin Institute of Technology in China have used Bayesian networks to standardize and automate a centuries-old method of diagnosing illness. Using pictures of the tongues of several hundred ill and healthy patients, they were able to build and test a network for diagnosing various illnesses. They estimate the prediction accuracy of their system to be up to 75.8%.

Download the paper (PDF)

Publicize Your Belief Network Success Story
Want to show off your applications of belief networks to others using them in their research? Send us your success stories. We would like to include at least one in each edition of BNet.News.

Send us your success story

BNet News is published by Charles River Analytics, the company that brings you BNet.Builder and BNet.EngineKit.  View our Privacy Policy