Step 1: Enable AI Evaluation for Subjective (TITO) Questions

1. Feature Overview

Think Exam now supports AI-based evaluation for Subjective Type Questions (Type In Type Out – TITO) within tests.

This feature enables automatic evaluation of candidate responses for subjective questions using AI technology, significantly reducing manual evaluation efforts and improving result processing time.

The feature is configurable and can be enabled by the Admin at the test level.

2. Key Benefits

• Faster evaluation of subjective responses
• Reduced dependency on manual evaluators
• Improved result turnaround time
• Consistent evaluation methodology
• Configurable evaluation strictness levels
• Scalable for large candidate volumes

3. How the Feature Works

Navigation

Go to:

Admin → Test Settings → Test Options

Enable the following setting:

• AI Subjective Evaluation Level
  

This setting activates AI-based evaluation for Subjective (TITO) questions in the test.

4. Enabling AI Evaluation While Creating or Editing a Test

While creating or editing a test, enable the option:

AI-Based Evaluation for Subjective (TITO) Questions

Once enabled, the system will evaluate candidate responses for subjective questions automatically based on the configured AI evaluation level.

Step 2: Select Evaluation Strictness

After enabling AI-based evaluation for Subjective (TITO) questions, the Admin can configure the evaluation strictness level based on the desired assessment approach.

The evaluation level determines how strictly candidate responses will be matched and evaluated by the AI engine.

 

Evaluation Levels

	Description
Lenient:	Flexible evaluation with broader acceptance of answers. Suitable when variations in wording or expression are acceptable.
Moderate:	Balanced evaluation approach with a moderate level of answer matching and flexibility.
Strict:	More stringent evaluation with tighter answer matching criteria. Suitable for highly precise or technical responses.

 

Configuration Path

Go to:

Admin → Test Settings → Test Options

Select the preferred value under:

AI Subjective Evaluation Level

The selected level will be applied to all eligible Subjective (TITO) questions within the test.

Configuration Path

Go to:

Admin → Test Settings → Test Options

Select the preferred value under:

AI Subjective Evaluation Level

The selected level will be applied to all eligible Subjective (TITO) questions within the test

4. AI Evaluation Logic

The AI evaluation process works based on whether a model solution has been provided for the Subjective (TITO) question.

Case 1: Model Solution Provided

If a Subjective (TITO) question contains a predefined model solution, the AI evaluates candidate responses based on:

• The provided model solution
• The selected AI evaluation strictness level

This helps ensure more accurate and structured response evaluation.

Case 2: Model Solution Not Provided

If no model solution is added to the question:

• The AI evaluates responses independently using contextual understanding
• The selected evaluation strictness level is still applied during evaluation

This allows the system to assess descriptive responses intelligently even without a predefined answer.

5. Automatic Evaluation Flow

Once AI-based evaluation is enabled for the test, the evaluation process is carried out automatically.

Evaluation Process

1. Candidate attempts the Subjective (TITO) question
2. Candidate submits the test
3. The system automatically:
• Evaluates the response using AI
• Assigns marks to the question based on the configured evaluation level
4. Regenerate the result to reflect the evaluated marks

6. Manual AI Evaluation (Post-Test Enablement)

If the AI evaluation setting is enabled after candidates have already submitted the test, the Admin can manually trigger the AI evaluation process.

Navigation

Go to:

Admin → Report → Subjective Report Evaluation → Click on Test Name → AI Evaluation → Regenerate Result

What Happens After Triggering AI Evaluation

Once initiated:

• The system automatically evaluates all pending Subjective (TITO) responses
• Marks are assigned based on the configured AI evaluation settings
• Updated results are generated accordingly