Abstract: Suicide is a significant global issue, with approximately 800,000 people taking their own lives every year. Detecting individuals at risk of suicide remains a challenging task, as highlighted in numerous suicide studies. However, with the widespread use of social media platforms, we have observed that individuals often express their suicidal thoughts or experiences in public on these networks. Therefore, it is crucial to identify people who may be prone to suicide at an early stage.
In this paper presents a novel approach for detecting suicidal content in social media platforms using natural language processing (NLP) and machine learning techniques. The proposed method combines keyword-based detection, sentiment analysis, and NLP-based approaches to identify posts that may indicate suicidal ideation. By analyzing the language and sentiment used in social media posts, the method aims to identify content that suggests a person may be at risk of suicide.
The method is developed by training it on a carefully curated dataset of labeled posts, which includes examples of both suicidal and non-suicidal content. Through rigorous evaluation using metrics such as precision, recall, and F1-score, the effectiveness of the proposed method is assessed. The results demonstrate that the method achieves a high level of accuracy in identifying suicidal content.
The implications of this research are significant. Social media platforms can incorporate the proposed method as an automated tool to flag potentially concerning content for further review. Trained mental health professionals can then examine the flagged posts and provide appropriate support and intervention to individuals in need. By leveraging this technology, timely interventions can be initiated to prevent suicides and offer assistance to those who may be at risk.
Overall, the proposed method offers a promising solution to the challenge of detecting suicidal content in social media platforms. By leveraging NLP and machine learning techniques, it provides a proactive approach to identify individuals who may be in distress, enabling timely intervention and potentially saving lives.
|
DOI:
10.17148/IJARCCE.2023.12676
