<![CDATA[Healthy mental health is a condition when minds are in a state of peace and calm and not disturbed, thus enabling us to enjoy our daily lives with respect for others. In Indonesia, there are quite a number of sufferers of mental health disorders, approximately 19 million people over the age of 15 experience mental and emotional disorders, both at mild to severe levels. These data show that the Indonesian state has not been able to properly address mental health problems and that the existence of a pandemic tends to increase sufferers of mental health disorders, which if left unchecked will have a negative impact. Based on this problem, the Circle application, that focuses on mental health using Android technology that supports self-help with several services, one of which is an article service. The article service in the Circle application requires a recommendation system that can recommend articles according to the mental health conditions experienced by users so that the articles are able to alleviate the mental health problems currently experienced by users. Topic Modeling is an approach to analyze a collection of text documents and group them into topics. Topic Modeling has several methods that can be applied in making topics, one of which is BERTopic. It is a technique that leverages Transformer and c-TF-IDF to create dense clusters, preserving keywords in topic descriptions while making topics easier to interpret. There are 3 important components of the BERTopic algorithm namely Document Embedding, Document Clustering, Topic Representation. This study uses Topic Modeling with the BERTopic method as the baseline for the mental health article recommendation system in the circle application.]]>
