AI Powered Answer Paper Grading System

Evaluation of descriptive and handwritten examination answers remains a challenging and time-intensive taskin academic institutions. Manual grading methods are proneto subjectivity, inconsistency, and delayed feedback. This paperpresents an AI Powered Answer Paper Grading System thatautomates the assessment of descriptive answers using OpticalCharacter Recognition (OCR), Handwritten Text Recognition(HTR), Natural Language Processing (NLP), and Machine Learning (ML). The system converts handwritten answer scriptsinto machine-readable text and performs semantic evaluationby comparing student responses with reference answers usingtransformer-based language models. Experimental observationsindicate that the proposed system significantly improves gradingconsistency, reduces evaluation time, and provides meaningfulfeedback. The solution aims to enhance fairness, scalability, andefficiency in academic assessment while reducing the workloadof educators.Index Terms—Automated grading, Handwritten text recognition, Natural language processing, Machine learning, Educationalassessment