NLP-Enhanced Intelligent Complaint Management System for Municipal Services

Abstract

Municipal complaint management systems traditionally rely on manual processes that are inefficient, error-prone, and lack real-time transparency. This paper presents an NLP-Enhanced Intelligent Complaint Management System for Municipal Services that leverages Natural Language Processing (NLP) and machine learning techniques to automate the classification, prioritization, and routing of citizen complaints. The proposed system processes complaints submitted via web platforms and applies text preprocessing, feature extraction using TF-IDF and word embeddings, complaint classification, sentiment analysis, and automated routing to the appropriate departments. The system provides officers with a dashboard to track and update complaint statuses, while citizens receive real-time notifications. Experimental evaluation demonstrates that the proposed system significantly reduces manual processing time, improves classification accuracy, and enhances citizen satisfaction. The work draws inspiration from the Traffic Urgency Model (TUM) proposed for Indonesian traffic complaints and extends the concept to a broader municipal services domain using a robust NLP pipeline.

Introduction

Across all the texts, a common theme is the use of modern AI and automated systems to replace slow, manual processes in critical real-world domains such as healthcare, security, governance, and public services.

One text discusses an intelligent organ donation platform that digitizes donor registration, centralizes data, and enables faster matching between donors and recipients to reduce delays and save lives. Another focuses on AI-based myocardial infarction detection using ECG signals, where CNN models, explainability techniques like Grad-CAM, and a web application are used to improve early diagnosis, especially in areas lacking cardiologists.

A facial recognition system for missing persons is also described, using deep learning models (such as CNNs, FaceNet, and DeepFace) to match uploaded images against databases, significantly reducing search time and improving identification accuracy. Similarly, a secure electronic voting system is proposed that uses dual biometric authentication (fingerprint and facial recognition), AI-based liveness detection, and fraud prevention mechanisms to ensure secure, transparent elections.

Another study examines browser extension security under Chrome’s Manifest V3, arguing that even with stricter permissions, side-channel attacks can still leak user data through timing behavior, storage contention, and service worker activity, highlighting ongoing privacy risks.

A municipal complaint management system is also presented, using NLP techniques (TF-IDF, sentiment analysis, classification models) to automatically categorize complaints, assess urgency, route them to departments, and track resolution with transparency and real-time updates.

Finally, a summary of an NLP-based complaint system emphasizes improving municipal services by automating complaint handling, prioritizing urgent issues, and reducing delays through machine learning and language processing.

Conclusion

This paper presented an NLP-Enhanced Intelligent Complaint Management System for Municipal Services that automates the classification, prioritization, routing, and tracking of citizen complaints using machine learning and NLP techniques. The system eliminates the inefficiencies of manual complaint management, ensures urgent complaints are handled first, and provides real-time transparency to citizens through status notifications. The key contributions of this work are: (1) a complete NLP pipeline for municipal complaint processing including preprocessing, TF-IDF feature extraction, classification, and sentiment-based urgency assignment; (2) automated complaint routing to municipal departments; (3) a citizen-officer interaction platform with real-time status tracking; and (4) a keyword search module for complaint pattern analysis. Future work will focus on expanding the complaint dataset to include diverse municipal service categories, implementing real-time deep learning models for improved classification accuracy, and integrating a mobile application interface for enhanced citizen accessibility. Additionally, dynamic weighting mechanisms for urgency variables, inspired by the TUM framework [1], will be explored to further enhance the system\'s responsiveness to contextual changes.

References

[1] B. R. Lidiawaty, E. Suryani, and R. A. Vinarti, \"Traffic Urgency Model: A Novel Approach to Prioritize Complaint Texts Using Enhanced Named Entity Recognition for Bahasa Indonesia Cases,\" IEEE Access, vol. 13, pp. 57720–57733, 2025, doi: 10.1109/ACCESS.2025.3556423. [2] W. K. U. Nasrulloh, C. Hermanto, I. Jamaludin, and L. Emalia, \"Public Complaint Management Information System Based on Complaint Topic Classification Using the Naive Bayes Model,\" 2025. [3] P. Gupta, O. P. Ijardar, A. Jadhav, and V. Saheb, \"AI-Based Solution to Enable Ease of Grievance Lodging and Tracking for Citizens Across Multiple Departments,\" 2025. [4] S. Chen, Y. Zhang, B. Song, X. Du, and M. Guizani, \"An Intelligent Government Complaint Prediction Approach,\" Big Data Research, vol. 30, Nov. 2022, Art. no. 100336. [5] C. Vairetti, I. Aránguiz, S. Maldonado, J. P. Karmy, and A. Leal, \"Analytics-Driven Complaint Prioritisation via Deep Learning and Multi-Criteria Decision-Making,\" European Journal of Operational Research, vol. 312, no. 3, pp. 1108–1118, Feb. 2024. [6] J. Blümel and M. Zaki, \"Comparative Analysis of Classical and Deep Learning-Based Natural Language Processing for Prioritizing Customer Complaints,\" in Proc. 55th Hawaii Int. Conf. Syst. Sci., 2022, pp. 1–16. [7] N. Kim and S. Hong, \"Automatic Classification of Citizen Requests for Transportation Using Deep Learning: Case Study from Boston City,\" Information Processing and Management, vol. 58, no. 1, Jan. 2021, Art. no. 102410. [8] K. Cabello-Solorzano et al., \"The Impact of Data Normalization on the Accuracy of Machine Learning Algorithms: A Comparative Analysis,\" in Proc. Int. Conf. Soft Comput. Models Ind. Environ. Appl., Jan. 2023, pp. 344–353.

Copyright

Copyright © 2026 Javith Kausar M, Mohamed Aathil M, Mohamed Asjath M, Mohamed Anas M, Mrs.G.BAKIYA JOSEPHA. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.