Sentiment Analysis and Text Classification for Depression Detection

Depression is an illness that can harm someone's life. However, many people still do not know they are having depression and tend to express their feelings through text or social media. Thus, text-based detection could help in identifying the early of illness. Therefore, research aims build a identify possible cues based on Bahasa Malaysia text. The data, form text, has been collected from depressed healthy via google form. There three questions asked which “Apakah kenangan manis yang anda ingat?”, rutin harian anda?” keadaan membuatkan stress?” obtained 172, 169 170 responses for each question respectively. All datasets stored CSV file. Using Python, TF-IDF was extracted as feature pipeline into several classifier models such Random Forest, Multinomial Naïve Bayes, Logistic Regression. results were presented using classification metrics confusion matrix, accuracy, F1-score. Also, another method conducted sentiment techniques Vader Text Blob onto whether depressive falls under negative vice versa. percentage differences determined between actual compared sentiment. From experiment, highest score achieved by AdaBoost Classifier with 0.66-F1 score. best model chosen be utilized Graphical User Interface (GUI).