Mental Health Sentiment Analysis on Social Media TikTok with the Naïve Bayes Algorithm

Mental health disorders can affect a person's emotions and behavior and can impact their health and daily activities. The most serious consequence of poor mental health is death by suicide. The problem in this research is about mental health which can be analyzed using the field of natural language processing with the science of sentiment analysis on the developing TikTok platform. Content contained on the TikTok platform will have comments made by fellow users, then these comments are collected for sentiment analysis. This study utilizes a classification algorithm, namely naïve Bayes. The dataset obtained is preprocessed by text first, then the process from TF-IDF will be used to be able to see the appearance of words in the data. The data used in this study is 6300 data. The results of this study are accuracy as measured by a metric evaluation which produces 80.95%. The sentiment in this study focuses on positive and negative sentiments. With visualization, words that often appear are tired, sick, tired, and hurt in negative comments. Meanwhile, the words that often appear in positive comments are able, passionate, strong, and happy.