Skripsi
PENGEMBANGAN CHATBOT UNTUK KESEHATAN MENTAL DI ERA GENERASI Z MENGGUNAKAN FRAMEWORK RASA
Generation Z is the largest demographic group in Indonesia and is increasingly vulnerable to mental health challenges. This study develops and evaluates a mental health chatbot using the Rasa framework and Natural Language Processing (NLP) to provide initial support for Generation Z. The chatbot was developed following the Rational Unified Process (RUP) methodology and integrated with the Telegram platform. A dataset comprising 58 mental health–related intents was used, and intent recognition was performed using the Dual Intent and Entity Transformer (DIET) classifier. System performance was evaluated through baseline and modified testing scenarios. The results show that the chatbot achieved strong classification performance, with an F1-score of 96.7%, precision of 96.8%, and recall of 97.2%. These findings indicate that the proposed chatbot can effectively support preliminary mental health assistance for Generation Z.
No other version available