Research On Item Selection Optimization Methods Based On Improved Simulated Annealing Algorithm

The development of educational evaluation has produced a new organizational form along with the continuous leap of computer and network technology.The new form of educational evaluation based on computer is alive: computer adaptive testing(CAT).Computer adaptive testing is a product of the full coupling of modern educational technology,psychometric theory and computer.It is based on the item response theory.Its guiding ideology is that the computer imitates the behavior of human intelligence assessment,by providing the subject with suitable fit.Test the ability level of the item,and observe the subject's reflection on the item to estimate the ability level of the subject.Therefore,computer adaptive testing can truly achieve the goal of "being human policy."The item selection method is the most important part of computer adaptive testing.Its quality is not only related to the safety and practicability of the item question bank in the system,but also related to the accuracy of the evaluation of the real ability level of the testee.Therefore,this paper takes the efficiency of item selection and the speed of item selection in the item selection method as the research objectives.The main research work includes the following points:Firstly,the item response theory and computer adaptive test are introduced and summarized.At the same time,the current status of popular item selection methods at home and abroad is analyzed and summarized.According to the item selection efficiency and item selection speed existing in the current item selection method,based on the item selection method based on the previous simulated annealing algorithm,the genetic algorithm and the simulated annealing algorithm are combined to propose a hybrid based Heuristic item selection optimization method: genetic-simulated annealing algorithm(TIS-GASA).Through the computer adaptive test system realized by ourselves,the experimental research on the genetic-simulated annealing item selection method proposed in this paper is carried out,which proves the high efficiency of the new item selection method.The innovation of this paper is to combine genetic algorithm and simulated annealing algorithm to form a new heuristic item selection method.Firstly,the algorithm searches for the current optimal solution by genetic algorithm,and uses the current optimal solution as the initial solution in the simulated annealing stage.And then further explore the optimal solution.While retaining the strong global search ability of the genetic algorithm,the algorithm also absorbs the powerful local search ability of the simulated annealing algorithm,avoiding the local optimal situation.Finally,the results of the experiment show that the genetic-simulated annealing algorithm has greatly improved the efficiency of item selection,and can provide the test questions with the most suitable of information to the subjects,and also has a significant improvement in the speed of the topic selection.