Smart Group Volume System Research Based On Genetic Algorithm

With the development of computer-aided instruction and artificial intelligence technology, more and more experts focus their attention to the study on the system of intelligent test paper. The establishment of the test database based on computer and realization of auto composing test paper is the important measure to make the examination more scientific and standard, then it is a key way to separate the examination from teaching. It is very helpful for the improvement of teaching quality. The efficiency and quality of automatic composing test paper is determined entirely by the design of test papers database and the selection of the test questions algorithm. It refers to not only the establishment of test paper automatic composing mathematical model, but also the application of the related algorithm in test paper composition system. The purpose of this paper is how to make the test system meet the examination requirements quickly and efficiently.First of all, based on the analysis of a large number of references home and abroad, the description of the intelligent test paper composition and the mathematic model is studied. All properties such as average difficulty, the proportion of various questions, teaching requirements, examination time, test scores are simplified, and the major methods are analyzed and evaluated. Based on that, several major key difficult problems and techniques in automatic composing test paper system are forwarded in this paper.Generation of the test paper is complex combinatorial optimization problem. Genetic algorithm lays a new foundation for development of the test paper composition system. Considering several main factors, such as encoding method, initialization colony, adjustment function, selection operator, crossover operator, variation operator, etc. And then it was applied into the actual intelligent test paper composition system.Based on a detailed system needs analysis, test paper composition system was designed. Test composition paper system composed of the user management module, subjects management module, test management module and the test paper management module four parts, of which bank management module is the core of the system, including the entry questions, query questions, modify questions three sub-modules. Intelligent test paper systems was studied and implemented in this paper can manage all kinds of questions be of different subjects, knowledge points, types. It can generate automatically test paper according to different requires. Success rate and efficiency in the test paper can meet the actual needs.