As online education increases, various education-related applications are being launched. However, it is difficult to present various types of problems in the existing education-related applications, and it is provided based on problems that have been registered in advance. Therefore, when new educational materials are added, it is difficult to separately create and provide new problems based on the contents of the materials. To improve these limitations, this study proposes a method that can help personal learning by automatically creating various types of problems through analysis of the provided lecture materials. In other words, if lecture materials are given, the document area is seperated from the text area of the lecture materials, and various types of problems are generated and provided through the problem generation algorithm after extracting candidate objects necessary for problem generation through the BERT Korean named entity recognizer(NER). It also proposes ways to increase the effectiveness of education through the management of incorrect answer notes.