<![CDATA[ Hibiscus flower extract (Hibiscus rosa-sinensis L.) has been widely studied and utilized in various fields. Nevertheless, a structured depiction of its utilization patterns based on existing scientific literature remains unavailable. This study seeks to systematically map and examine the primary focus areas of hibiscus flower extract usage by employing Word Cloud visualization alongside Naïve Bayes classification techniques. The core dataset consisted of research abstracts retrieved from Google Scholar, PubMed, and ScienceDirect over the past ten years. Additionally, data from social media platforms (Twitter/TikTok) were gathered as a supplementary source to capture public perceptions. All collected texts underwent preprocessing steps including stopword removal and stemming prior to analysis. The resulting Word Cloud highlighted the most prominent terms: "hibiscus flower," "extract," "antimicrobial," "natural dye," "cosmetics," "antioxidant," "antibacterial," "skin care," and "medicinal plant"—findings that were further supported by high classification probabilities derived from the Naïve Bayes model. When evaluating three Naïve Bayes kernels in combination with Word Cloud visualization, the Multinomial Naïve Bayes using a 90:10 training-to-testing data split achieved optimal performance, yielding an accuracy of 88.42%, precision of 87.91%, recall of 88.15%, and an F1-score of 88.03%. Moreover, the Word Cloud revealed that the "Health" category dominated academic discussions (42%), featuring keywords such as herbal, blood pressure, and antioxidant. This was followed by the "Research" category (35%) with terms like extract, test, and flavonoid, and the "Beauty" category (23%) focusing on skin, mask, and anti-aging—identifying a promising direction for future research.]]>
