Sistem Informasi Re-Actions pada Fitur Sentimen Data Aduan Publik dan Reporting

Abstract

An effective public complaint management system plays an essential role in enhancing governmental transparency, accountability, and responsiveness to public needs. This study aims to develop Re-Actions, a web-based information system designed to facilitate the structured collection, processing, and monitoring of public complaints. The system integrates an automatic sentiment analysis feature to identify emotional tendencies or public opinions from complaint texts using a machine learning model. The software was developed using the Scrum methodology, enabling an iterative and adaptive development process aligned with user requirements. The sentiment analysis model was built using the Support Vector Machine (SVM) algorithm and trained on 2,756 public complaint records obtained from the archival data of the Layanan Aspirasi Kotak Saran Anda (LAKSA) system of Tangerang City. Experimental results show that the sentiment analysis model achieved an accuracy of 79%, indicating a reliable capability in classifying public complaints into positive, negative, and neutral categories. This level of accuracy is consistent with previous studies on machine learning-based sentiment analysis in public service domains, which generally report performance within the 70%–80% range, depending on data characteristics and applied methods [3], [12], [13]. Furthermore, the system was evaluated using Black-box Testing to verify functional correctness and User Acceptance Testing (UAT) to assess usability and user satisfaction. All core system features operated as expected, and the UAT results indicated a user satisfaction rate of 92%, reflecting a high level of system acceptance and consistency with similar information system evaluations in the public sector [16]. These findings demonstrate that integrating machine learning techniques into public complaint information systems can enhance information management effectiveness, accelerate data-driven decision-making, and support improvements in public service quality in Tangerang City.