<ul><li>Suicidal thoughts and behaviors are a growing societal concern, prompting the need for tools to detect suicidal risk early.</li><li>Researchers have developed a robust machine learning model using Reddit posts to classify them into four levels of suicide risk.</li><li>The model combines RoBERTa, TF-IDF, and PCA to enhance accuracy and reliability in assessing suicide risk severity.</li><li>Through experimentation, the hybrid model outperformed RoBERTa-only and traditional machine learning classifiers, achieving a best weighted $F_{1}$ score of 0.7512.</li></ul>

Detection of Suicidal Risk on Social Media: A Hybrid Model

Discover more