LLM-as-Judge in Education: A Curriculum-Grounded Marking Pipeline

2026-06-16

AI Summary

This study addresses the challenge of aligning large language model (LLM)-based automated scoring with official curriculum standards in high-stakes assessments. It proposes a configurable LLM-as-Judge scoring pipeline that maps test items to curriculum ontologies to identify corresponding topics, subtopics, and cognitive demands. The approach translates curricular intent into structured rubrics—specifying action verbs and performance-level descriptors—and leverages educational knowledge graphs and official guidelines to guide the LLM through staged, rubric-compliant scoring. Its key innovation lies in explicitly encoding curriculum standards into actionable curricular artifacts, thereby ensuring transparency, consistency, and traceability in scoring. Preliminary evaluation demonstrates strong agreement between system-generated scores and human raters, with justifications directly traceable to authoritative standards. The system has been integrated into an online learning platform and is now generating empirical usage data.

Abstract

Generative AI and large language models (LLMs) are increasingly applied to question generation and automated assessment. However, deploying LLMs in preparation for high-stakes exams requires more than prompt engineering; it demands software pipelines that systematically ground model outputs in authorised curriculum artefacts and marking guidelines issued by education authorities. This paper presents a curriculum-grounded, configurable LLM-as-Judge pipeline for question-level marking, co-developed with an industrial partner, to support exam preparation for university admission. The pipeline identifies the relevant topics, subtopics, and cognitive demand of a question, and assembles verifiable and authorised context to support LLM judgement. Curriculum intent is operationalised through concrete syllabus artefacts, including prescribed verbs and outcomes, performance band descriptors, glossary definitions, and marking-guideline principles. A staged LLM workflow is employed to first generate question-specific rubrics, capturing structured expectations of performance, and then derive and evaluate marking criteria used to allocate marks to student responses. This design improves consistency, transparency, and alignment with official marking practices. Preliminary evaluation shows that the proposed LLM-as-Judge pipeline delivers marking outcomes comparable to human tutors, while yielding justifications that are more traceable to authorised curriculum artefacts and marking standards. The pipeline has also been integrated into an online study platform, where early deployment data provide initial insights into operational usage and manual overrides.

Problem

• LLM-as-Judge
• automated assessment
• curriculum grounding
• marking guidelines
• high-stakes exams

Innovation

• LLM-as-Judge
• curriculum grounding
• automated assessment
• rubric generation
• marking pipeline