WSDM2021

Towards Ordinal Suicide Ideation Detection on Social Media

Ramit Sawhney 1 Saumya Gandhi 3 Harshit Joshi 2 Rajiv Shah 1
1IIIT Delhi, India
2Cluster Innovation Center, India
3Visvesvaraya National Institute Of Technology, India

The rising ubiquity of social media presents a platform for individuals to express suicide ideation, instead of traditional, formal clinical settings. While neural methods for assessing suicide risk on social media have shown promise, a crippling limitation of existing solutions is that they ignore the inherent ordinal nature across fine-grain levels of suicide risk.