AUTHENTIWRITE: A HANDWRITING RECOGNITION DEVICE FOR STUDENT VERIFICATION IN MODULAR DISTANCE LEARNING

The introduction of modular distance learning 
(MDL) presented major challenges regarding the verification 
of students’ handwritten output to ensure academic integrity 
and fairness of assessment. As part of this challenge, this 
research study was undertaken to design and develop 
AuthentiWrite, an artificial intelligence‑based handwriting 
recognition device that verifies student authorship through 
offline processing. Based on the developmental–experimental 
research design methodology and the V‑model framework 
methodology, the system integrates hardware components, 
namely the Raspberry Pi, camera module, and sensors, with 
machine learning techniques such as the Siamese neural 
network and support vector machine for the analysis and 
comparison of handwriting. A total of 30 respondents 
participated in a controlled system testing and evaluation 
process utilizing ISO/IEC 25010 standards. Calibration of the 
system took place to improve the performance of the system’s 
ability to distinguish genuine from forged handwriting. 
Results indicated that the average accuracy of the system was 
87.5%, the mean response time was 24.88 seconds, and the 
continuous operating error rate was 0%. In addition to these 
results, an overall rating of acceptance for functionality, 
reliability, efficiency, usability, maintainability, and 
portability was provided by the 30 respondents with a value 
of 4.36. It can be concluded that AuthentiWrite is a reliable 
solution for improving academic integrity and fair assessment 
within modular distance learning.