Mining Twitter Data to Improve Detection of Schizophrenia.

摘要: Individuals who suffer from schizophrenia comprise I percent of the United States population and are four times more likely to die suicide than general US population. Identification at-risk individuals with is challenging when they do not seek treatment. Microblogging platforms allow users share their thoughts emotions world in short snippets text. In this work, we leveraged large corpus Twitter posts machine-learning methodologies detect schizophrenia. Using features tweets such as emoticon use, posting time day, dictionary terms, trained, built, validated several machine learning models. Our support vector model achieved best performance 92% precision 71% recall on held-out test set. Additionally, built a web application that dynamically displays summary statistics between cohorts. This enables outreach undiagnosed individuals, improved physician diagnoses, destigmatization