Research On Intelligent Test Paper Composition System Based On Genetic Algorithm

With the rapid development and wide application of the Internet, the intellectual test paper composition system base on the WEB technology has become an important method in assisting instruction. Intellectual test paper composition algorithm is the key in the design and realization of intellectual test paper composition system. Traditional test paper composition algorithm like random method and backtracking, etc. can't satisfy the real need in the aspects of success rate and efficiency of the test paper. However, the genetic algorithm has the features of adaptive global optimization, intellectual search and nice convergence, etc, which can meet the requirement of intellectual paper test composition better. The thesis mainly made a research in the application of generic algorithm in intellectual paper test composition.First off all, the thesis makes a description of the intellectual paper test composition and the mathematic model in the first place, analyzes and evaluates the major methods of intellectual paper test composition, then put forward the theory and method of intellectual paper test composition through the means of generic algorithm. Secondly, targets at the problems of application of traditional generic algorithm in test paper composition system like early convergence and low searching efficiency in the evolution anaphase, the thesis puts forward a practical intellectual test paper composition through optimizing design of the key technology like test paper composition process, genetic operators, population size and iterative suspension condition, etc. Lastly, the design and realization process of the general structure, database design, each functional module and the major algorithms of the system is described in great detail, and some concrete examples of paper test composition are also given in the thesis.Intellectual paper test composition researched and realized in the thesis can administrate test questions of different classes, knowledge points and question types; and can automatically make paper test composition by generic algorithm according to different requirements of test paper composition. The method can satisfy the realistic needs in the successful rate and efficiency of test paper composition.