Preventing Hidden Information Leaks Using Author Attribution Methods and Neural Networks

This paper addresses the problem of hidden information leakage detection through the use of text steganography. Presented comparative research results show how to perform this task by detecting changes in user's writing styles using neural networks and various types of text features. The framework for hidden leakages detection based on discovering changes in the author's writing style with deep ne...