Year: 2026
Type: article
Abstract: This paper formalizes the RAG-based automated grading pipeline, defines evaluation metrics for retrieval relevance and rubric alignment, and demonstrates that graph-structured RAG outperforms flat vector retrieval on complex assessment tasks.
Author: Pakutnik, D. V.
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7197394551