A unified approach to suicide risk detection using text and voice

Suicide affects over 3000 Australians every year. Traditional forms of assessment have not succeeded in addressing this concerning trend. This PhD thesis by publication addresses a pressing need for alternative forms of assessment and detection. This PhD explores the use of text-based and voice signal-based analyses, that together hold significant promise in identifying suicide risk and psychological distress in callers to national telephone helpline counselling services. It is further argued that these forms of analysis cannot occur without first addressing key ethical considerations. Thus, this PhD thesis offers an ethical, efficient and real-time detection of suicide risk.