Abstract- Evaluating answer scripts manually in educational institutions takes a lot of time, effort, and can sometimes be subjective. This research introduces Auto Grade, an AI-driven system designed to automate the grading process. Using Natural Language Processing (NLP) and Optical Character Recognition (OCR), Auto Grade analyzes typed responses from scanned PDFs. The system first extracts text through OCR, cleans it to remove noise, and then applies deep learning models like BERT and RoBERTa to compare student responses with model answers based on meaning. A scoring algorithm evaluates responses by considering content relevance, coherence, and completeness while also providing detailed feedback to help students learn better. Experiments show that Auto Grade significantly reduces grading time while ensuring consistency and fairness. The system is scalable, minimizes bias, and improves efficiency, making it a strong alternative to manual grading. Future improvements will focus on enhancing OCR accuracy, supporting multiple languages, and refining NLP models for specific academic fields.
Keywords :
Automated Grading, NLP, OCR, Deep Learning, Machine Learning
|
DOI:
10.17148/IJARCCE.2025.14579
