<![CDATA[The rapid advancement of information technology has encouraged the public to actively share reviews through digital platforms such as Google Maps. These reviews are not only informative but also reflect real user opinions and experiences regarding places or institutions, including higher education institutions. This study aims to analyze the main topics and sentiment classification contained in Google Maps reviews related to Universitas Pamulang. The approach used in this research combines two main methods. First, topic modeling is conducted using BERTopic, a modern technique based on transformer embeddings and HDBSCAN clustering algorithms, which can capture the semantic context of text more deeply. Second, sentiment analysis is performed using a lexicon-based approach, applying an Indonesian sentiment lexicon to efficiently identify the polarity of opinions without requiring model training.The data analyzed were collected through web scraping of relevant public reviews on Google Maps across four Universitas Pamulang locations: Central Campus, Viktor Campus, Witanaharja Campus, and Unpam Serang. The analysis revealed several dominant topics such as academic services, campus facilities, and bureaucracy. The majority of sentiments identified were neutral to positive, although negative opinions were also found in certain aspects. These findings are expected to serve as strategic input for the university to enhance service quality and strengthen its institutional image in the digital landscape.]]>
