<![CDATA[The development of online lectures in Indonesia has encouraged universities to develop more targeted branding strategies, particularly in increasing brand awareness and digital engagement. Public perceptions reflected in open digital sources, particularly online news portals, are important data sources for understanding how online lectures are perceived by the public. This study aims to analyze public opinion on online lectures in Indonesia using web scraping, text mining, TF-IDF Vectorizer, and the K-Means Clustering algorithm. Initial data was obtained through web scraping using several Indonesian keywords related to online lectures and resulted in 108 text documents. After duplicate removal, noise filtering, and text preprocessing, the number of documents used in the analysis was reduced to 60 documents. The analysis stages included text cleaning, normalization, feature extraction using TF-IDF, determining the optimal number of clusters using the Elbow Method and Silhouette Score, and clustering using K-Means. The results showed that the optimal number of clusters was 4 clusters. The formed clusters represent the themes of online lecture academic activities, social and cultural dimensions, distance education and policy, as well as economic, regulatory, and technological aspects. The TF-IDF matrix yielded 1669 features, while the final evaluation yielded a Silhouette Score of 0.0345. PCA visualization showed overlap between clusters, indicating that public perception of online lectures is multidimensional. The conclusion of this study shows that the TF-IDF and K-Means approaches can be used to map public opinion and develop more contextual, data-driven, and relevant online lecture branding recommendations for universities.]]>
