Abstract
Depression is one of the mental disorders that are often experienced by a person in daily life. Social media platforms is a new thing as an alternative to tell stories and express current feelings by people today. Twitter is one of the social media that is often used to express feelings and opinions through tweets posts, including tweets that contain hate speech which indirectly shows symptoms of depressive disorder through statements uploaded. It also requires modeling that can recognize users with the potential to experience depression so that they can get initial treatment. This can be implemented using the BiLSTM (Bidirectional Long Short-Term Memory) method and the Word2Vec feature. It can be concluded that the dimensional size of the large feature word2vec, LSTM, and Conv1d layers influenced the model in detecting depression which can be seen in the testing accuracy and F-1 score according to the split data used.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have