Manual generate of test papers is a difficult assignment for instructors, particularly inside a brief timeframe outline. It requires a large amount of work and time to accomplish a standard nature of the test papers. question paper necessarily to consider several items such as difficulty level marks, numerical as well as theoretical contents of the paper, weightage of questions, and repetition of exam questions in terms of their characteristics is one of the most important problems in generating the test. The scientific contribution of this research is a proposed framework based on combine between genetic algorithms and unsupervised learning methods (k-means) to generate a test paper according to Bloom's criteria, where the k-means method collects each group of questions with the same features in one group. where address the problem of the repetition of questions that have the same characteristics. In the exam paper, based on the previous step, it will be easy for the genetic algorithm to choose the best exam according to the six levels bloom's. The framework consists of several phases .in the first phase, generate the questions bank in the second phase, the questions were divided into six groups that were similar in their properties using the k means method, in the third phase, used a random sorting function to randomly arrange each group to ensure that the questions were not repeated when the initial population was created. Used A question bank of 800 questions including all types of questions.

Automated Exam Questions Generator Bloom’s Taxonomy Genetic Algorithm k-means