Deep Learning-Based Complaining Task Distribution towards Smart City

The absence of an automated online system to report complaints, along with the challenges faced by city corporations in handling them, poses significant hurdles. Additionally, there is a lack of resources for higher authorities to monitor the city corporations’ progress effectively. The primary objective of this research is to develop two Android applications aimed at automatically categorizing complaints. This was achieved using a deep learning model developed on the Teachable Machine platform, trained on a dataset comprising 5494 images of municipal issues. This study considered four categories of complaints, with each class comprising approximately 1500 images. Furthermore, measures were implemented to filter out fake complaints and ensure the accurate classification of images that were initially misclassified. The user application facilitates convenient reporting of complaints by capturing images of municipal issues, while the authority application delivers categorized complaints along with location and status updates. Higher authorities can oversee the progress of municipal operations, leading to increased transparency, and efficiency and fostering nationwide smart city development.