Proactive Detection and Intervention of Suicidal Ideation on Social Media

Mae'r digwyddiad hwn wedi dod i ben

Professor Dongsong Zhang is visiting us from University of Maryland has offered to give a talk entitled above to our staff and students on Wednesday 6 June at 1400h in C/2.07. He is hoping to collaborate with us through writing research grants and/or articles.

Abstract: Suicidal ideation, also known as suicidal thoughts, is extremely detrimental and dangerous and thus is considered as psychiatric emergency that requires immediate assistance from mental health professionals. Current intervention methods are passive and reactive because they require individuals with suicidal ideation to take the initiative, which is neither timely nor effectively. In view that people with suicidal ideation increasingly share their thoughts and emotions online, social media have become a viable source for timely detection and intervention of suicidal ideation. This study proposes, designs, deploys, and  assesses SOP3, a transformative and integrative approach to social media based suicidal ideation detection and intervention. Two key novel design artifacts of SOP3 include using a deep neural network model for effective detection of suicidal ideation from microblogs and a proactive, participatory, and problem-solving based intervention via a social media platform. We empirically tested those design artifacts through an analysis of real-world microblog posts collected from Sina Weibo and through an online field experiment with individuals who were detected with suicidal ideation. The results show that the proposed deep neural network models outperformed traditional classification models by a wide margin. The findings from both quantitative and qualitative feedback of the field experiment demonstrated users’ acceptance and positive perceptions of SOP3, which is critical for realizing its real-world impacts. The proposed research artifacts and findings of this study have significant research implications and broad practical impacts for improving the mental health of the public.

Bio: Porfessor Dongsong Zhang, Department of Information Systems at the University of Maryland, Baltimore County, USA, received his Ph.D. in Management Information Systems from the University of Arizona in 2002. His current research interests include mobile HCI, computer-mediated collaboration and communication, online deception detection, text mining, and health IT. He has published over 140 research articles in journals and conference proceedings, including ACM Transactions on Accessible Computing, Communications of the ACM (CACM), IEEE Trans on Knowledge and Data Engineering (TKDE), IEEE Trans on Software Engineering, IEEE Trans on Multimedia, IEEE Transactions on Systems, Man, and Cybernetics, Journal of the American Society for Information Science and Technology,  among others. He has received a dozen research grants and awards from U.S. National Science Foundation (NSF), National Institute of Health (NIH), National Natural Science Foundation of China, and Google Inc.