Transforming Handwritten Assessments with AI Grading and Analytics in PenScan

The increasing workload of educators, particularly in grading handwritten assessments, underscores the need for automated solutions that enhance efficiency while maintaining accuracy. This study introduces PenScan, an AI-powered application designed to automate the grading process for multiple-choice, true/false, and identification exams using Optical Character Recognition (OCR) technology. PenScan leverages Azure OCR to digitize handwritten responses, enabling automated scoring and item analysis. The system incorporates difficulty and discrimination indices to evaluate assessment quality and features a review mechanism allowing students to request grade revisions. The application's performance was assessed by evaluating handwriting recognition accuracy, grading reliability, and the effectiveness of its feedback system. Experimental results indicate that PenScan achieves an average OCR accuracy of 93% in recognizing handwritten responses and a grading accuracy of 85%. A usability survey conducted among educators and students showed a high satisfaction rate highlighting its effectiveness in streamlining assessment tasks while maintaining transparency in grading. PenScan demonstrates strong potential in reducing the manual effort required for grading, improving overall efficiency, and enhancing data-driven decision-making in education. This study contributes to the advancement of AI-assisted assessment tools, offering valuable insights into their integration and impact in academic settings.