Avaliação de ações preventivas de riscos utilizando teoria de decisão e redes de Petri coloridas

Abstract

Risk management in software projects involves the definition of actions to prevent risks identified for the project in order to minimize their effects or eliminate them.The definition of preventive actions, and especially to assess their efficacy in eliminating a risk is not a trivial task. The objective of this research is to identify and propose a method for evaluation of preventive actions to mitigate or eliminate risks in software projects. This assessment is based on supporting a quantitative analysis driven Decision Theory and modeled and simulated by means of colored Petri nets. The choice of theme reveals the importance of predicting the impact and efficacy of preventive measures in software projects, anticipating their possible outcomes and enhancing their uses. The development of this research resulted in three distinct stages of study, mutually complementary and performed in different periods: a) define an approach to model and simulate processes that were widely accepted by the scientific community, b) identify a theoretical basis that was able to establish a criterion to support the decision making process and therefore evaluate the impact of preventive risk in software development projects; c) evaluating the results of simulation based on the modeling of preventive risk using the previously established theoretical basis. The study is an exploratory, descriptive and analytical, combined with documentary analysis of literature sources, from documents and information from the literature. The proposed method consists of introducing a formal step in the evaluation process of preventive risk.The simulation with colored Petri nets, aided by the Theory of Decision by the Bayes Theorem, and has made the process more understandable, provided a more effective participation by experts involved, and allow formal mathematical representation coupled to mechanisms analysis to inspect risks adapted processes.