Intelligent Test Model-based Reasoning And Genetic Algorithms And Systems To Achieve

With the further development of computer and network technology, computer-based education has become an important part of modern higher education, and computer examination systems as an important component of computer-aided instruction are paid more and more attention. In the examination systems, the quality of automatic test papers directly affects on the true evaluation of the level of examinees and the fairness of the examination. Therefore, the research of algorithms and models of the intelligent test papers has also become an important topic of computer-aided education.The paper overviews the background of test papers, analyzes theories and technology of test papers and finds that at present, most of the intelligent test paper systems rely on Genetic Algorithm(GA) to solve multi-objective constrained optimized problems. If the test papers require more parameters, it will have more difficulty and GA will be easy to form an immature convergence, furthermore, the active participation of users must be ignored in the course of test papers and the practicality of test papers can be lowered. In the light of the above-mentioned problems, it proposes an intelligent test paper model based on reasoning and GA by comparing the features of various intelligent test paper algorithms, the key points of this paper as follows:1) It analyzes associated strategy of the targets of parameters system in the space constrained by multi-objection to achieve the purpose of dimensionality reduction, which can effectively prevent premature of GA test papers;2) The introduction of reasoning allows users to engage in pre-selection test questions, and it can form pre-selected Item Bank to reduce the range of the parameters and narrow the scope of test papers, which can increase the practicality and flexibility of test papers;3) GA optimizes test papers at the basis of the pre-selected Item Bank and the system can generate multiple sets of test papers in five-dimensional goal space constrained to meet user requirements;4) Through the analysis of test papers, on the one hand, the system can examine the quality of the whole test papers to update the knowledge of test papers in accordance with the requirements in Case Base; on the other hand, it can examine the quality of the test questions and use it as feedback information to dynamically update the Item Bank to improve the construction of Item Bank..