<![CDATA[This study investigates public sentiment toward gamification features in e-commerce platforms, focusing on user opinions expressed on social media platform X. As gamification strategies like Shopee Candy and Tokopedia Quiz become more common, understanding user perceptions is crucial for improving user experience. This research adopts a hybrid approach using Convolutional Neural Network (CNN) for sentiment classification and Latent Dirichlet Allocation (LDA) for topic modeling. A total of 765 user comments were collected using Tweet Harvest and processed through standard text preprocessing techniques. The CNN model achieved strong performance, with 81% accuracy, 86% precision, 89% recall, and an F1-score of 88%. LDA analysis revealed key terms in each sentiment group—positive sentiments centered on words like “menang” and “hibur,” while negative sentiments included “scam” and “capithaha.” The results demonstrate the effectiveness of combining CNN and LDA for analyzing sentiment and extracting dominant themes in informal online discussions. These insights can guide e-commerce platforms in refining gamification strategies that align with user expectations, helping to enhance engagement while addressing user concerns.]]>
Copyrights © 2025
