Data Quality Matters: Suicide Intention Detection on Social Media Posts Using RoBERTa-CNN.

Area of Science:

• Computational linguistics
• Artificial intelligence
• Mental health technology

Background:

• Suicide is a major global health issue requiring advanced detection methods.
• Online platforms like Reddit contain valuable data for identifying at-risk individuals.
• Existing methods may not fully capture the nuances of suicidal ideation expressed in text.

Purpose of the Study:

• To propose and evaluate a novel deep learning model for identifying suicidal intentions in Reddit posts.
• To leverage the strengths of RoBERTa and CNN for enhanced text analysis.
• To investigate the impact of data quality on model performance for suicide detection.

Main Methods:

• Utilized the RoBERTa-CNN deep learning model, combining RoBERTa for semantic understanding and CNN for pattern recognition.
• Trained and evaluated the model on the Suicide and Depression Detection dataset.
• Implemented data cleaning techniques, including manual cleaning and OpenAI API, to improve text data quality.

Main Results:

• The RoBERTa-CNN model achieved a mean accuracy of 98% with a standard deviation of 0.0009.
• Demonstrated that data quality significantly influences the robustness and performance of the detection model.
• Data preprocessing steps were crucial for optimizing model training.

Conclusions:

• The RoBERTa-CNN model shows significant promise for the early detection of suicidal intentions in online text.
• High-quality, cleaned data is essential for developing effective AI-driven mental health tools.
• This approach offers a scalable solution for monitoring and intervention in online mental health contexts.