Automated Answer Script Evaluation and Grading System

Abstract- Manual grading of student response scripts for subjective examinations are a time-consuming and tedious process for educators, often leading to inconsistencies and fatigue during evaluation. This paper presents an automated system employing machine learning techniques to evaluate studentresponses against predefined standards, aiming to improve grading accuracy, consistency, and fairness andreduce human effort for the same. The primary objectives are enhancing grading effectiveness for teachers byautomating script examination to reduce time and effort while allowing more feedback and better interactionand focused learning; improving accuracy by mitigating subjectivity of manual grading; and promotingequality by ensuring all students are evaluated by the same criteria, minimizing bias. The system features anintuitive interface for uploading PDF answer scripts, automatically scanning content and employingalgorithms to assess responses. Graded results are provided in downloadable files for analysis and recordkeeping. By providing objective, consistent assessments, the automated approach streamlines grading,enhances transparency, and contributes to an equitable, effective learning environment. Leveraging techniqueslike Optical Character Recognition (OCR), text similarity measures, machine learning models, naturallanguage processing (NLP), and semantic analysis, the proposed system aims to provide an accurate andefficient automated evaluation of student responses.