Improving Municipal Responsiveness Through AI-powered Image Analysis in E-Government

Abstract

Integration of Machine Learning (ML) techniques into public administration marks a new and transformative era for e-government systems. While traditionally e-government studies focused on text-based interactions, this one explores the innovative application of ML for image analysis, an approach that enables governments to address citizen petitions more efficiently. By using image classification and object detection algorithms, the model proposed in this article supports public institutions in identifying and quickly responding to evidence submitted by citizens in picture formats, such as infrastructure issues, environmental concerns or other urban issues that citizens might face. The research also highlights the Jevons Paradox as a critical factor, wherein increased efficiency from the citizen side (especially using mobile platforms and apps) may generate higher demand which should lead to scalable and robust solutions. Using as a case study a Romanian municipality that provided datasets of citizen-submitted images, the author analysed and proved that ML can improve the accuracy and responsiveness of public institutions. The findings suggest that adopting ML for e-petition systems can not only enhance citizen participation but also speed up administrative processes, paving the way for more transparent and effective governance. This study contributes to the discourse on e-government 3.0 by showing the potential of Artificial Intelligence (AI) to transform public service delivery, ensuring sustainable (and scalable) solutions for the growing demands of modern urban governance.