AI Rubrics

The OpenEduCat AI Rubrics suite provides an end-to-end intelligent evaluation ecosystem for academic institutions. By integrating AI agents with OpenEduCat’s Assignment and Grading systems, the suite covers the entire rubric lifecycle—from rapid criteria generation and deep pedagogical design to automated, rubric-aligned submission grading.

Suite Architecture

The AI Rubrics framework comprises two core capabilities:

1. AI Rubric Generator (openeducat_ai_rubric_generator): An intelligent wizard to design, generate, and iteratively refine standardized scoring rubrics based on grade level, assessment types (Formative, Summative, Authentic, etc.), and learning objectives.

2. AI Rubrics Assessment (openeducat_ai_rubrics_assessment): Automates assignment grading by analyzing student-submitted PDF documents against rubric matrices, computing scores, and providing line-by-line feedback with teacher override controls.

Key Features

• Automated Rubric Generation: Create comprehensive rubrics with criteria descriptions, performance tiers, and balanced point weights in seconds.

• Pedagogical Alignment: Support for multiple assessment types (Formative, Summative, Diagnostic, Authentic, Peer, Project-Based) and rubric formats (Analytic, Holistic, Single-Point, etc.).

• Iterative Refinement: Interactive Regenerate & Refine workflow allowing educators to adjust specific criteria using natural language prompts before saving.

• Direct Template Creation: One-click export to OpenEduCat Rubric Templates (op.rubric.template) for immediate assignment attachment.

• Automated PDF Submission Grading: Extracts text directly from student PDF submissions and evaluates them against active assignment rubric criteria.

• Interactive Rubric Matrix Widget: Visual AI evaluation button embedded directly in the grading interface, auto-populating level selections and feedback.

AI Rubric Generator

Configuration

Navigate to Settings ‣ AI Settings (or General Settings ‣ AI Agents):

• Locate the Rubric Generator section.

• Enable Deep Analysis for Rubric Generation.

• In the Rubric Deep Agent field, select the active AI Agent used for generating and refining rubrics.

Workflow: Generating & Refining Rubrics

1. Launching the Rubric Generator Wizard

• Navigate Assignments ‣ Configuration ‣ Generate Rubrics (AI).

2. Configuring Rubric Parameters

Fill in the initial generation criteria:

• Subject / Task: Enter the assignment title or prompt (e.g., Persuasive Essay on Renewable Energy or Physics Lab Experiment Analysis).

• Grade Level: Select from Elementary (K-5), Middle School (6-8), High School (9-12), Undergraduate, or Graduate / Professional.

• Rubric Type: Choose Points (numerical weights) or Percentage.

• Assessment Type: Select the pedagogical context (Formative, Summative, Authentic, Project-Based, etc.).

• Rubric Format: Select the structure (Analytic, Holistic, Single-Point, Developmental, or Checklist).

• Learning Objectives: Optionally add target learning competencies (e.g., Critical Thinking, Evidence Synthesis, APA Citation Format).

3. Executing Generation & Preview

• Click Deep Analysis Generation.

• The AI Agent computes balanced weights (automatically normalized to 100 total points) and renders a live interactive HTML table preview showing all criteria, level descriptions, and point allocations.

4. Iterative Refinement

• If any criterion requires adjustment, type modification instructions in the Adjustment Notes field (e.g., « Make the research methodology criterion more rigorous » or « Add a specific row for presentation formatting »).

• Click Regenerate & Refine to update the rubric structure while maintaining overall context.

5. Saving to Rubric Templates

• When satisfied with the generated matrix, click Save Rubric Template.

• The system automatically generates an active OpenEduCat Rubric Template record with all nested criteria (op.rubric.element) and level descriptors (op.rubric.element.level).

AI Rubrics Assessment & Grading

Configuration

In Settings ‣ AI Settings: * Locate the Rubrics Assessment section. * In the Rubrics Assessment Agent field, select the default active AI Agent responsible for analyzing student PDF submissions and scoring rubric matrices.

Workflow: Evaluating Submissions

1. Attaching Rubrics to Assignments

1. Open an assignment in Assignments ‣ Assignments.

2. Under the Rubrics configuration, select the generated Rubric Template.

3. Under the Rubrics AI Configuration tab, confirm the active Rubrics AI Agent.

4. Publish the assignment to students.

2. Evaluating Student Submissions

1. Open a student’s submission in Assignments ‣ Submissions.

2. Open the grading view containing the student’s submitted PDF document and the embedded rubric matrix.

3. Click the AI Grade button (indicated by the robot icon on the rubric matrix).

4. The system executes the background analysis workflow: * Extracts text from the uploaded PDF submission. * Compiles the assignment’s active rubric elements and level definitions. * Prompts the configured Rubric Assessment Agent. * Normalizes and parses the structured response.

5. The Rubric Matrix widget automatically updates in real time: * Selects the appropriate radio button level for each rubric criterion. * Populates criterion-specific feedback text below each element. * Computes the calculated Final Score and overall assessment summary.

3. Teacher Review & Finalization

• Teachers retain full pedagogical control and can manually adjust any level selection, edit comments, or override the final score before saving the submission.

Agent Library Templates

The AI Rubrics suite includes pre-built templates in the Agent Library:

• Rubric Generator (Deep Analysis): Standard template configured for structured JSON generation, automatic point balancing, and multi-format alignment.

• Rubrics Assessment Agent: Pre-configured with safety guardrails and structured evaluation schemas to evaluate PDF submissions against rubric levels.