Efficient Electronic Document Access Control Management using Natural Language Processing

Electronic Document Access Control and Management has been challenging as large scale organizations and various governments are transitioning towards an electronic media platform. The electronic medium includes the majority of the electronic documents that are utilized for their ease in sharing and storage. The increasing number of documents that are being stored are becoming highly difficult to manually store and manage. The risk of sensitive documents being accessed by an unauthorized person is also remaining as a looming threat that needs to be eliminated. There has been a considerable amount of research has been performed for enabling an effective Document Management System. But most of them fall short of the expectations due to their inability to accurately perform automatic access control. Therefore, this paper outlines an effective and secure technique for the maintenance of an access control mechanism through a machine learning approach. The presented technique implements Term Frequency - Inverse Document Frequency (TF-IDF), Entropy Evaluation along with the Decision tree to achieve an accurate and reliable Document Management System. The automatic electronic document access control relieves the manual calibration of the Document Management system significantly. The proposed research experiment demonstrates a comparison with the traditional document management system, where the proposed EDACM system yields a Root Mean Square Error score of 1.48, which is significantly less and achieved an accuracy of 85.71%, denotes a better fit model. Thus, the proposed approach exclusively boosts the performance of Document Access Control Management.