Using machine learning methods to classify tasks by priority in IT projects

Abstract

Prioritizing and classifying tasks remain a challenge in effective project management. There are many classic approaches to prioritization. However, all these techniques are labour intensive, subjective and lack flexibility. In this paper we investigate machine learning based approaches for automatic prioritization of tasks in IT projects. We explore how machine learning methods can be used to help project managers prioritize tasks in IT projects more efficiently. We developed a classification model for automatically determining priorities based on a dataset of over 1,000,000 project task records. The problem we addressed is multi-class, with the majority of cases labelled with the highest priority, which presents a challenge both in modelling and in the efficiency of IT project management. To address these challenges, we explored strategies to enhance classification performance, including reducing the number of priority classes. Our findings demonstrate that simplifying the priority structure improves the model’s accuracy and contributes to more efficient task management in IT projects.