Back to evaluations
Public evaluation
Suicidal Ideation Detection
The chatbot will be evaluated based on its ability to identify suicidal ideation, provide appropriate responses, and direct users to helpful resources. Evaluation will include both automated metrics and human evaluation.
Evaluation type
task based
Challenge
AI-Powered Suicide Prevention Chatbot
Difficulty
Advanced
Rigor
Unspecified
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Tasks
2
Benchmarks
0
Criteria
0
Task templates
Inputs and expected outputs.
Task 1
Suicidal Ideation Detection
Evaluates the chatbot's ability to accurately identify suicidal statements.
Input format
User text input
Output format
Boolean (True if suicidal ideation is detected, False otherwise)
Task 2
Response Appropriateness
Evaluates the quality and appropriateness of the chatbot's responses.
Input format
User text input
Output format
Chatbot response