With the development of automation in all industries, E-Learning automation: automatically generate E-Learning contents gives E-Learning teachers new opportunities to implement effective management of E-Learning systems. And many eLearning professionals already use E-Learning templates and online asset libraries to create cost effective eLearning courses. Moreover, all E-Learning Automation methods have one thing in common—algorithms: used to control various aspects of e-learning course creation. Additionally, another important thing is online exams which are used to get an immediate feedback from learners. This research focuses on automation generation of E-Learning contents, which can be mainly divided into three parts. Firstly, automatically summarize relevant documents using natural language processing and deep learning techniques. Secondly, detecting keywords in generated summaries and then delete them from input summaries. Finally, rearrange output results and fill them into E-Learning system.