Microsoft MSBNx

Category Intelligent Software>Bayesian Network Systems/Tools

Abstract MSBNx (Bayesian Network Editor and Tool Kit) is a component-based Windows application that supports the creation, manipulation and evaluation of Bayesian probability models, created at Microsoft Research.

Each model is represented as a graph or diagram.

The random variables are shown as ellipses, called nodes, and the conditional dependencies are shown as arrows, or directed arcs, between variables.

At the present time, MSBNx only supports discrete distributions for its model variables. Models are saved to and loaded from disk-based text files in an Extensible Markup Language (XML)-based format. MSBNx supports simultaneous viewing and evaluation of multiple models.

File Formats -- The only document type in MSBNx is the model network document. Each saved document represents one Bayesian belief network and includes its prior probabilities and the properties associated with the model, including all of its variables, or nodes.

MSBNx supports two (2) distinct file formats:

1) The XBN format is XML-based and is the default file format used by MSBNX.

2) The DSC format is a legacy text format used within Microsoft Research.

The XBN format is intended to replace the legacy DSC format.

Both of these file formats are basically text and can be edited or generated by such tools as EMACS, Notepad or WordPad.

Bayesian Networks -- Bayesian Networks are useful for diagnosis and troubleshooting.

Recommendations --

When doing diagnosis and troubleshooting, MSBNx can recommend what evidence to gather next.

If you give MSBNx cost information, it does a cost-benefit analysis. If No cost information is available, MSBNx makes recommendations based on the Value of Information (VOI).

Assessment --

MSBNx tries to make it easy for you to specify your probabilities for a Bayesian Network:

1) With the Standard Assessment Tool, you can specify full and causally independent probability distributions.

2) With the Asymmetric Assessment Tool, you can avoid specifying redundant probabilities.

3) If you have sufficient data and use 'machine learning' tools to create Bayesian Networks, you can use MSBNx to edit and evaluate the results.

Use It in Programs --

MSBNx is fully component based. Its most important component is MSBN3, an ActiveX Dynamic-Link Library (DLL).

MSBN3 offers an extensive Component Object Model (COM)-based Application Programming Interface (API) for editing and evaluating Bayesian Networks. You'll find MSBN3 especially easy to use from COM- friendly languages such as Visual Basic and JScript.

MSBNx also includes graphical components, for example, both the Standard Assessment and Asymmetric Assessment tools are ActiveX controls and can be used in other applications.

Also, you can extend the editing and evaluation abilities of MSBNx by creating add-ins. For example, MSBNx ships with an (undocumented) add-in for editing and evaluating Hidden Markov Models (HMMs).

Note: Inference on Hidden Markov Models is exact but inefficient for large models.

Comprehensive Help Pages --

1) Help pages for both the MSBNx editor and the MSBN3 API are on-line.

2) The manuals are also installed as Windows Help documents.

3) While editing with MSBNx or while programming with MSBN3, pressing F1 will bring up help for the topic on which you are currently working.

The MSBN3 API Help includes overviews of loading models, running models, and other common programming tasks.

Dynamic Properties --

Either interactively or programmatically, you can define and attach property values to models and nodes. For example, you can define a property type called “DatabaseKey” with string values.

Then on each node in a model, you can attach a string value to use for looking up more information related to the node in a database. Model and nodes can have any number of properties. Property values can be strings, reals, arrays of string or reals, or enumerated types.

System Requirements

The application and its components run on Windows 98, Windows 2000, Windows XP and Vista.

Manufacturer

Manufacturer Web Site Microsoft MSBNx

Price Contact manufacturer.

G6G Abstract Number 20207

G6G Manufacturer Number 101785