Enhancing Aspect-Based Sentiment Analysis through Data Labeling Classification on Student Reviews Using a Text Sampling Approach

Abstract

This study aims to enhance aspect-based sentiment analysis through data labeling classification on student reviews using a text sampling approach. We employ the K-Nearest Neighbor (k-NN) method to group similar reviews based on specific aspects. The dataset utilized consists of student reviews regarding their experiences collected from online questionnaires in private universities in Indonesia. This research successfully improves the understanding of sentiments expressed in student reviews by employing the text sampling approach and data labeling classification. The k-NN method yields more accurate predictions in identifying sentiments related to various aspects of the reviews. The practical implication of this research is the enhancement of aspect-based sentiment analysis on student reviews.