Auto-Generating Test Paper Algorithm With Multi-Directional Evolution Genetic Algorithm

With the continuous development of computer technology, the application of CAT (Computer Assist Teaching) has become more and more widespread. The Test Questions Base System is one of them. Whether at home or abroad, the theories of the test questions base system is developing continuously, which have great effect on the improvement of teaching methods and quality. In recent years, the enlarging recruitment of colleges and universities, the increase of students, the open window to the public of high mathematics classes and the use of grade teaching model by colleges anduniversities-----all make much more request to the examine methods and the selectedtopics quality and the examine times. Genetic Algorithm has made a number of achievements in all fields, which offers untapped potential for solving those complex problems such as NP and fitness system. This thesis introduces the development status of the test questions base system, compares the organization of test paper automatically by use of the test questions base system with the traditional one by hand and also lists the existent questions about the forming by test questions base system. Furthermore, this thesis put forward a new intellectual algorithm model which combines the genetic algorithm frame with the normal distribution. This model has made some alterations and optimizing to the primary population and select and evolution strategy of genetic algorithm based on the test paper system. This paper makes a detail introduction to the test paper generating algorithm of the intellectual system of item database.