Implementation of Essence Practice into Bayesian Networks

Abstract

Modern project management in software engineering still lacks a commonly accepted applicable formal model of projects that can be used for accumulating experience or for developing applied optimization methods that can potentially help IT companies to reduce risks and deliver software in a more efficient way. The Essence language which was developed by SEMAT initiative can potentially work as such a formal model for software projects, but current Essence practitioners mostly focus on methodology work for describing different approaches to perform tasks for software projects. In this work, we propose a way to develop a prototype of a decision support system based on the Essence kernel and language in combination with an applied optimization model. In order to do this, we firstly design how to include Essence practice in an applied math model and then modify a Bayesian network that finds false-positive manager mistakes. As an example of implementation of the achieved results, we present the current state of the plugin for the software project management system Redmine that uses our approach to help managers of projects.