Facilitating and Expediting the Development of Systems that Handle Uncertainty through PRoModel

Rosa A. López-Landa, Marcos A. Bernal, and Julieta Noguez

Keywords

graphical user interfaces, intelligent tutoring systems, model-driven development, probabilistic relational models, software architectures, web-based software engineering

Abstract

This paper introduces PRoModel, a Web tool that expedites the development of applications that handle uncertainty. PRoModel is based on a service-oriented, model-driven and multilayered architecture that provides flexible and reusable applications. The graphical user interface developed in the presentation layer facilitates the design of Probabilistic Relational Models (PRMs), allowing the inclusion of (1) the entities of the problem and their relationships from an Entity-Relationship (ER) model, and (2) the random variables and conditional dependence associations from some Probabilistic Graphical Models (PGMs), such as Bayesian networks. Once the user completes the design of the PRM, the logic layer of PRoModel transforms the model to generate a functional Web application prototype that facilitates the management of the domain data and provides the initial PGM. To demonstrate the use of PRoModel we have generated an Intelligent Tutoring System (ITS) for training operators of power plants. Initial tests show that PRoModel expedites and facilitates the development of Web applications with uncertainty models, providing significant savings in time and effort without requiring the user to be an expert in systems development.

Important Links:



Go Back