Kairós

Abstract

This paper presents a computational model entitled Kairós for prediction and recommendation in project schedules. The model uses context prediction mechanisms based on task data and projects stored during its execution. The recommendations are made to the manager in a proactive manner, considering best practices in project management and learning with the approval or rejection of each recommendation. A prototype was implemented based on the proposed model, and through it, an evaluation was carried out using simulated use cases with real data from a large company. The results showed that the model was able to predict with precision of 93% if a task would be completed with delay, with 87% accuracy.