Suicidal Electra
S
Suicidal Electra
Developed by gohjiayi
A text classification model based on the ELECTRA architecture, designed to detect suicidal tendencies in text.
Downloads 50
Release Time : 3/2/2022
Model Overview
This model predicts whether a text contains suicidal tendencies (1 indicates presence, 0 indicates absence), suitable for mental health monitoring and early warning.
Model Features
High accuracy
Achieves an accuracy of 97.92% on the test set, demonstrating excellent performance.
Efficient training
Utilizes single-round training (1 epoch) and a small batch size (6), saving computational resources.
Balanced dataset
Training data includes 232,074 text records evenly distributed between suicidal and non-suicidal categories.
Model Capabilities
Suicidal tendency detection
Text classification
Use Cases
Mental health
Social media monitoring
Monitor user posts on social media platforms to identify potential suicidal tendencies.
Helps mental health professionals promptly identify high-risk individuals.
Psychological counseling assistance
Assists psychological counselors in assessing clients' suicide risk.
Provides objective text analysis results to support professional judgment.
Featured Recommended AI Models
Š 2025AIbase