Research Of Dynamic Optimization Of Test Difficulty Coefficient In Item Bank

With the further development of internet and educational reform, various education forms such as online learning have emerged. Online learning is prominent in these forms. The pertinence of questions problem is particularly important. At present, the determination method of the questions difficulty coefficient in online learning system is mainly for the artificial estimation and adjustment.The process of artificial estimation cannot be avoided and there is no way to improve.However, domestic and foreign scholars have engaged in a large number of theory and technology research for many years.The use of the adjustment is more and more accurate and scientific.In the development of the era of Big Data, the determination of questions difficulty coefficient still has inevitable shortcomings:how to shorten the adjustment range of user. When users using the test may have problems such as the difficulty coefficient is not accurate.How to reduce user problems will become a big challenge of the questions difficulty coefficient optimization algorithm. How to use the existing test data of the new questions make adjustment difficulty coefficient is also another big problem we need to think about.In view of the above two problems, this paper research the existing question bank system, include the questions difficulty coefficient optimization algorithm and the application, mainly completed the following work:1) The author systemizes algorithm researches of questions difficulty coefficient recently, and uses the classical test theory as the theoretical basis to dynamic adjustment the question bank.2) The study the effect of adding sample library on the original item difficulty coefficient, and optimizes the corresponding algorithm, which solves the problem of reducing the adjustment range of users.3) The author improved the algorithm flow, using the existing questions for a lot of practice, test and adjust the existing database system.4) The author verify the application scope and feasibility of questions difficulty coefficient algorithm through practical application.