Deep Learning - Driven RPA Bot for Automated Evaluation of Handwritten Examination Papers Using Gen AI

摘要

The incorporation of Robotic Process Automation (RPA) and deep learning in the educational evaluation system brings a strong, automated solution for assessing handwritten exam assessments. Educational institutions have exponentially started using RPA systems for monotonous activities like data storage and management, but the capacity extends far beyond these activities. This paper gives an automated solution for evaluating handwritten examination papers in educational institutions by combining advanced deep learning models and RPA (Robotic Process Automation). Using CNNs (Convolutional Neural Networks) and BiLSTM (Bidirectional Long Short-Term Memory) networks, the system converts handwritten responses into machine-readable text, trained to various handwriting styles. For grading, a Gen-AI model is used to assess the answers recognized by CNN and BiLSTM. This provides precise grading that aligns closely with human correction. RPA does end-to-end automation, from scanning and processing papers through deep learning models for text recognition to grading and data management thus reducing human effort, enhancing consistency, saving time and improving efficiency in the evaluation process. This technique streamlines examination, allowing teachers and professors to focus more on personalized instruction and academic development.