Identification of Suicidal Despair using Convolutional Long-Short-Term-Memory Network and Natural Language Processing (NLP) techniques
Disciplines
Computer Sciences | Data Science | Psychiatry and Psychology
Abstract (300 words maximum)
Suicide is a serious and complex issue affecting many individuals and communities worldwide. It can be a tragic outcome of various factors, such as mental illness, social isolation, and difficult life circumstances. Therefore, it is essential to recognize the warning signs of suicide and seek help for anyone struggling with suicidal thoughts or behaviors. Following this research, we utilize an in-depth overview of the suicidality of 500 anonymous Reddit user posts, which store the information in datasets via python. Within the dataset, the data is sorted by categories such as user, post, and label using the Columbia Suicide Severity Rating Scale (C-SSRS). For example, the label ranging from 0-4, with four being at the highest risk for suicide behavior and 0 being no suicidal ideations or plans, were observed in the Reddit post. Through our research, we wish to provide a safe environment for those at risk of suicide and help prevent these impactful actions by detecting these suicide events using natural language-based machine learning techniques. We envision developing a Convolutional Long-Short-Term-Memory (CNN + LSTM) network to identify the suicidal risk category from Reddit Post. Furthermore, by providing such a program across platforms, suicidality can be decreased by a substantial amount, saving lives worldwide.
Academic department under which the project should be listed
CCSE - Computer Science
Primary Investigator (PI) Name
Md Abdullah Al Hafiz Khan
Identification of Suicidal Despair using Convolutional Long-Short-Term-Memory Network and Natural Language Processing (NLP) techniques
Suicide is a serious and complex issue affecting many individuals and communities worldwide. It can be a tragic outcome of various factors, such as mental illness, social isolation, and difficult life circumstances. Therefore, it is essential to recognize the warning signs of suicide and seek help for anyone struggling with suicidal thoughts or behaviors. Following this research, we utilize an in-depth overview of the suicidality of 500 anonymous Reddit user posts, which store the information in datasets via python. Within the dataset, the data is sorted by categories such as user, post, and label using the Columbia Suicide Severity Rating Scale (C-SSRS). For example, the label ranging from 0-4, with four being at the highest risk for suicide behavior and 0 being no suicidal ideations or plans, were observed in the Reddit post. Through our research, we wish to provide a safe environment for those at risk of suicide and help prevent these impactful actions by detecting these suicide events using natural language-based machine learning techniques. We envision developing a Convolutional Long-Short-Term-Memory (CNN + LSTM) network to identify the suicidal risk category from Reddit Post. Furthermore, by providing such a program across platforms, suicidality can be decreased by a substantial amount, saving lives worldwide.