Mental health patients on Twitter
In recent decades, mental illnesses have become a global challenge.
In Australia, one in five people experience mental illness every year. Although these illnesses are not incurable, most people affected do not seek treatment.The goal for this project is to investigate the language pattern of people with mental illness and build a model that can recognise signs of mental illness in a person's posts on social media.
In this project, we used tweets from Twitter as training data. We then used various machine learning algorithms to identify features that could differentiate user posts that showed signs of mental illness from those that didn't. We present a model that can receive a tweet from Twitter and decide whether it contain signs of mental illness. Overall, this significantly supports people with potential mental health issues by identifying signs of mental illness early, with the aim of greatly reducing the negative impact of these illnesses.
Dung Anh Hoang