Informatics, TU Vienna

DBL Interactive – A new online tool for Bringing Bayesian Networks to the world

Bayesian networks are a form of artificial intelligence, and over the past decade, their popularity has increased across a number of disciplines.

Abstract

Bayesian networks are a form of artificial intelligence, and over the past decade, their popularity has increased across a number of disciplines included computer science, engineering, medicine, business and finance, forensics, robotics and natural resource management. They can be used to integrate qualitative and quantitative information; to integrate expert knowledge, empirical data and outputs from other models in decision making; to explicitly accomodate uncertainty in deision making; and to conduct rapid scenario and sensitivity analysis. Bayesian networks are also a low cost means of developing deccision support tools and they can be developed and used overthe internet, thus supporting eResearch and collaborative learning.

This presentation will provide an introduction to Bayesian networks and demonstrate DBL Interactive that is specifically designed to support the collaborative development of decision support tools using Bayesian networks. Examples will cover its application to natural resource management. DBL Interactive, however, might be used in any area and even by people with no modeling background.

Biography

Since completing a PhD at UQ in 1998 I have worked for the Queensland Department of Natural Resources, the University of Melbourne and the Victorian Department of Natural Resources and Environment, developing decision support tools for natural resource management in collaboration with Catchment Management Groups and Authorities. My current area of research is in the participatory development of decision support tools with Regional NRM Bodies and land managers using Bayesian Belief Network Models. This is being conducted through the CRC for Tropical Savannas Management.