Automatic question generation is a popular field for many researchers because it is vital in reducing the time and effort of educators in many fields. Still, the main focus is on the English language, and the tools for question generation in Arabic still need to be developed. This paper seeks to overcome this limitation by proposing a method to automate or semi-automate the task of question generation for the Arabic language focused on history topics, using natural language processing's rule-based approach and employing the Arabic language's structural characteristics. Our method successfully generated fill-in-the-blanks MCQ with distractors, and True/False questions with the correct answer, offering the capability of automatic grading without any reliance on external sources or pre-trained models. The proposed method was tested on an Arabic article that contained historical information about the Sumerian civilization and used to generate 100 fill-in-the-blanks MCQ and 100 True/False questions. A manual evaluation by history experts was performed on the generated questions. The generated MCQ achieved (82%) of acceptable questions, and the True/False questions achieved (70%). The findings in this research provided a promising approach toward automating the question generation for the Arabic language while minimizing the required computational resources.