AI-Based Evaluation of Handwritten Scripts Using OCR and NLP : A Deep Learning Approach to Automated Assessment

Abstrak

Manual evaluation of handwritten answer sheets is often time-consuming, error-prone, and subject to human bias. This paper presents a novel AI-driven framework for automating the evaluation of handwritten descriptive answer scripts using a combination of Optical Character Recognition (OCR), Natural Language Processing (NLP), and Deep Learning. The proposed system first digitizes handwritten scripts using high-resolution scanning, followed by accurate text extraction with OCR. The extracted text is then analyzed using NLP techniques—such as tokenization, POS tagging, and named entity recognition—to understand the semantic structure and context of responses. A deep learning model based on Convolutional Recurrent Neural Networks (CRNN) further refines the evaluation process by assessing content relevance against predefined keywords and patterns. The system ensures a more objective, scalable, and efficient grading process. By integrating computer vision with advanced language models, this approach significantly reduces examiner workload, enhances grading fairness, and opens new avenues for intelligent educational assessment.

Identifier

• DOI: https://doi.org/10.32628/ijsrset2513812
• OpenAlex: https://openalex.org/W4415587919