Introduction > QGeNIe

QGeNIe is an interactive development environment for rapid creation of qualitative causal models in uncertain domains. These models represent propositions by means of nodes in an acyclic directed graph. These nodes are always propositional and take two possible values: True and False. The colors of these nodes represent the degrees of truth of the propositions. Mathematically speaking, the colors represent the probability of the state True (or False - it is the users' choice). While QGeNIe users can define the color scale, the default is a range between red and green, representing undesirable and desirable states respectively. QGeNIe allows for an interactive exploration of the models, examining the effects of observations and manipulations of individual variables.

There are two major applications of QGeNIe:

1.A standalone system for rapid creation of causal models, useful in all kinds of strategic planning problems, where problems are complex enough to be a challenge for an unaided human mind and, at the same time, too complex to model by means of fully specified, precise quantitative models. For sufficiently complex problems, it is a challenge for an unaided human mind to predict effects of various actions. Typically, in addition to obvious consequences of a decision, there will be indirect pathways through which actions may propagate through the system and lead to surprising effects. QGeNIe is a tool for capturing the knowledge and intuitions of decision makers and focusing group discussion on calculating the global effects of various decision options. QGeNIe offers what can be called an instant gratification system in the sense of showing interactively the effects of observations and manipulations.

2.A system for generating simple first-cut version of quantitative probabilistic models. Models developed by means of QGeNIe can be exported to GeNIe and refined into fully quantitative models.

An early version of QGeNIe was created and developed at the Decision Systems Laboratory, University of Pittsburgh between 1995 and 2015. In 2015, we created a company, BayesFusion, LLC, and acquired a license for QGeNIe from the University of Pittsburgh. Continuing the tradition of the Decision Systems Laboratory, we are making it available free of charge to the academic community for research and teaching use in order to promote decision-theoretic methods in decision support systems. QGeNIe has been tested extensively in many teaching, research, and commercial environments. We are continuously improving it and are interested in user comments. We encourage the users of QGeNIe to let us know about encountered problems and possible suggestions.

GeNIe's name and its uncommon capitalization originates from the name Graphical Network Interface, given to the original simple graphical user interface to SMILE, our library of classes for graphical probabilistic and decision-theoretic models. Its successor was GeNIe, SMILE's GUI. QGeNIe is a simplified, qualitative graphical user interface to SMILE.

QGeNIe allows for building models of any size and complexity, limited only by the capacity of the operating memory of your computer. QGeNIe is a modeling environment. Models developed using QGeNIe can be exported to GeNIe or embedded into any applications and run on any computing platform, using SMILE, which is fully portable.

QGeNIe, GeNIe and SMILE have been originally developed to be major teaching and research tools in academic environments and have been used at hundreds if not thousands of universities world-wide. Most research conducted at the Decision Systems Laboratory, University of Pittsburgh, found its way into both programs. Because of their versatility and reliability, QGeNIe, GeNIe and SMILE have become incredibly popular and became de facto standards in academia, while being embraced by many government, military, and commercial users.

The strongest element of GeNIe (and indirectly of QGeNIe), one that distinguishes it from a large number of other graphical modeling tools, is its user interface. We have paid a lot of attention to it and it shows. While developing decision-theoretic models takes typically an enormous amount of time, GeNIe cuts the effort by orders of magnitude and it will lead to a fast return of the investment in its licensing fees. SMILE is not far behind and belongs to the easiest to learn and use, most reliable, and fastest libraries for graphical models.