Design of an academic document forgery detection system

Abstract Advances in digital technology and communication are reducing the popularity of physical or paper-based documents, particularly in the official academic field. Many academic institutions have shifted from paper-based academic documents to electronic formats. However, with the availability of document-editing and photo-editing software, an adversary can illegally modify original academic documents to suit their needs. Therefore, it is essential that receivers of these digital academic documents must verify the authenticity and detect the forgery of the documents. In this study, a cryptographic hash function, which is concerned with data integrity, is used to detect modifications of digital academic documents. Experiments were conducted to evaluate the correctness and speed of academic document verification. The proposed system was compared with other conventional methods such as digital signature-based, convolutional neural network, and blockchain-based methods. Our study shows that the proposed system can efficiently detect forged digital academic documents with 100% accuracy and a verification speed of 0.352 ms, which is considerably faster than the current digital signature-based, convolutional neural network and blockchain-based technologies. The proposed cryptographic hash function-based method provides another way to help detect forged academic documents.