Design of an Auto Evaluation Model for Subjective Answers Using Natural Language Processing and Machine Learning Techniques

Manual grading in educational assessment presents inherent challenges, motivating the development of automated systems. This paper proposes an Automated Evaluation System leveraging natural language processing (NLP) and machine learning techniques. The system’s architecture integrates NLP algorithms and machine learning models such as Gradient Boosting to predict scores based on semantic similarities with expected answers. Empirical results showcase the system’s proficiency in accurately evaluating student responses. Additionally, the system incorporates Optical Character Recognition (OCR) for handwritten paper checking, enhancing its versatility and applicability. In conclusion, QualiScore offers a robust solution to the complexities of manual grading, facilitating efficient and objective assessment practices in educational institutions.