<![CDATA[Information technology has significantly influenced various sectors, including education, where it is increasingly used to evaluate and improve teaching systems. However, traditional evaluation methods often yield superficial feedback that lacks specific, actionable insights. This study aims to deeply analyze the sentiment of the teaching system evaluation at Kampung Inggris Bandung EPLC and present the findings intuitively through visual analytics. Data were gathered from student feedback during the 79th learning period, collected via Google Forms between January 7 and January 26, 2025. A total of 2,590 raw data entries were preprocessed through rigorous Natural Language Processing (NLP) techniques and analyzed using the K-Nearest Neighbor (KNN) algorithm paired with Term Frequency-Inverse Document Frequency (TF-IDF) feature extraction. The classification resulted in 1,171 positive sentiments and 111 negative sentiments. The model evaluation demonstrated high effectiveness, achieving an accuracy of 90.66%, precision of 82.19%, recall of 90.66%, and an F1-score of 86.22%. Furthermore, data visualization using Wordclouds successfully mapped student response patterns across various class categories (Rookie, Speaking, Grammar, and TOEFL), highlighting institutional strengths in engaging teaching methods and pinpointing specific facility-related areas for improvement.]]>
Copyrights © 2026
