Analisis Pemodelan Topik Saran Pengguna Jalan Tol dengan Latent Dirichlet Allocation (LDA)

Service and safety for toll road users are two crucial aspects in the operation of toll roads. This research aims to evaluate and analyze toll road users feedback after receiving assistance by officers and determine the dimensions of Service Quality that have been fulfilled from the distribution of topics obtained. The method applied is Latent Dirichlet Allocation (LDA), a topic modeling technique that can analyze thematic patterns in text collections. By applying LDA to the dataset of toll road user suggestions, this research seeks to uncover the main topics related to improving service quality in toll road operations. Through this analysis process, the company can recognize common problems that are often emphasized by toll road users and understand their needs and expectations for improving toll road service quality. This research shows that the company has fulfilled four service-related dimensions namely reliability, assurance, responsiveness, and tangibles. This shows that the company has carried out the duties and responsibilities of service to road users well. With this research, the company is expected to be able to improve service-related performance in order to maintain positive sentiment for the company's performance in the eyes of the community.