| With the rapid development of the modern network technology and the reformation of teaching work, more and more education institutions need to complete the inspection work by a swift and smart option. In this situation, the on-line examination system which is based on the computer technology arised. Through analyzing the current research of on-line examination system, the paper adopted the. net development platform, with SQL Server 2008 database for support, designed and developed an on-line examination system based on the B/S mode.In this paper, the function, database and sets of system algorithm were analyzed, designed and realized. According to the requirement analysis, three kinds of roles which are administrator, instructor and student were designed. Different functions within this system were specified for the proper authorized users. Management and maintenance of the related data were done by the administrator. Management of the examination paper and marking subjective items were finished by the teacher. Student is the subject of attending on-line examination.Through the comparison of several common strategies of grouping, the genetic algorithm which has a good convergence was chose as the strategies of intelligent fixing test paper. In this paper, the genetic algorithm was improved from four aspects:coding scheme, calculation method of crossover probability, calculation method of mutation probability and selection operator. The paper used the real-coded schema, adopting paper numbers as gene values. Adaptive method was used to set crossover probability and mutation probability. Elite selection strategy combined with wheel wager choice was used as selection operator of the genetic algorithm.The system was tested, by manual methods, automated testing tools. Test results showed that system runs stably and responds quickly. Feasibility and efficiency of the improved genetic algorithm were verified by the experiment. |
PDF Full Text Request