High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Automated Assignment Evaluator Using Retrieval-Augmented Generation

Author(s):

Gayatri Mudkanna , Zeal Polytechnic, Pune; Raj Mohite, Zeal Polytechnic, Pune; Shritej Waikar, Zeal Polytechnic, Pune; Shravani Pande, Zeal Polytechnic, Pune; Sushmia Choure, Zeal Polytechnic, Pune

Keywords:

Retrieval-Augmented Generation, Automated Assignment Evaluation, Large Language Model, Semantic Similarity, Rubric-Based Grading, Educational Technology

Abstract

Automated assignment evaluation can reduce the time required to assess student answers, but evaluating answers written in different words remains a challenge. This paper proposes an Automated Assignment Evaluator using Retrieval-Augmented Generation (RAG). The proposed system retrieves relevant information from approved study materials and supplies it to a large language model along with the question, reference answer, and marking rubric. The model then generates a suggested score and feedback based on the retrieved evidence and the rubric. The workflow includes answer extraction, document retrieval, rubric-based evaluation, feedback generation, and teacher review. The system is intended to support educators rather than replace their final judgement. Its performance should be assessed by comparing its scores with human-assigned marks and measuring scoring error, agreement, evaluation time, and feedback usefulness. As the system is proposed, experimental results are not claimed in this paper.

Other Details

Paper ID: IJSRDV14I90027
Published in: Volume : 14, Issue : 9
Publication Date: 01/12/2026
Page(s): 47-48

Article Preview

Download Article