Advances in machine learning in suicide prediction on online social platforms

This article systematically reviews the research results related to the machine learning based suicide ideation prediction on social networking platforms, so as to provide references for group and individual suicide prediction. This article will address the current states (issues of algorithm accuracy and efficiency, privacy leakage and stigma) and limitations of machine learning based suicide prediction on different platforms (light blogging, acquaintance social platforms, forums, picture and video sharing applications and clinical databases).