Febriyanti, Eni (2025) Analisis Sentimen Berbasis Aspek Terhadap Pelayanan Rumah Sakit Menggunakan Metode K-nearest Neighbors Dan Naive Bayes. Other thesis, Universitas Islam Riau.

Text 203510417.pdf - Submitted Version Restricted to Registered users only Download (9MB) | Request a copy

Abstract

Hospitals are organizations that provide professional medical services with adequate facilities, supported by the attitudes and behaviors of medical personnel that reflect professionalism and high commitment, which are the main factors in maintaining patient trust and loyalty. Google Maps provides location, route and direction information, as well as user reviews that can help understand public sentiment towards a place. However, inconsistencies between ratings and comments indicate the need for further analysis, especially on hospital service reviews in Pekanbaru. This research focuses on topic modeling and aspect-based sentiment analysis of hospital reviews in Pekanbaru, including Hermina, Awal Bros, Sansani, Santamaria, Syafira, Prima, Eka Hospital, Annisa, Arifin Ahmad, and Aulia. Topic modeling is carried out with the aim of obtaining topics or aspects on comments using Latent Dirichlet Allocation which are evaluated and analyzed using the elbow method based on coherence and perplexity values. Next is to perform aspect-based sentiment analysis based on the aspects identified in topic modeling using 2 machine learning approaches, namely Naïve Bayes and K-Nearest Neighbor which are evaluated by comparing accuracy values. The aspects identified in the topic modeling include administration and service aspects. The results of aspect-based sentiment analysis show Naïve Bayes is the best classification model in all tasks with an accuracy of 84.6% in the aspect classification task and 91.9% in the administration aspect and 96.7% in the service aspect in the aspect-based sentiment classification task.

Actions (login required)

View Item