The Design And Accomplishment Of Intelligent Test Paper Composition System Based On Chaos Genetic Algorithm

The rapid development of computer technology and artificial intelligence brings social working efficiency to be soared.All parts of educational fields have influence with the same tendency without exception.In teaching activities,the evaluation of teaching effect is examination-based all the time.Computer technology and artificial intelligence merge into Test Paper which is beneficial for improving working efficiency and declining labor intensity.In order to make the examination paper which constitutes by automatic Test Paper more scientific and objective,more brilliant Test Paper algorithm are needed.This dissertation illustrates the basic theory and principle of Test Paper.Furthermore,it analyzes which basic principle and limitation should be followed by Test Paper system.In order to make Test Paper System acquire the best pattern of Test Paper as much as possible,the dissertation is based on Chaos in Genetic Algorithms(CGA)to avoid random Test Paper algorithm of non-optimization problem.At the same time,it solves the problem when putting genetic algorithm directly into application of optimization problem,traditional genetic algorithm easily traps into local optimization and operation time is long or so.This algorithm speeds up global optimization ability and decrease the proportion of local optimization.Base on CGA is to bring chaos in genetic algorithms,to execute Multi Segment Integer Number Coding from initial population of traditional genetic algorithm.This encoding method is more straightforward and easy to understand.It is convenient for operation,and it could do Chaos Optimization separately by selecting operation,interlace operation and mutation operation.To effectively avoid original algorithm brings a series of problems.To some extent,it promotes execution efficiency for algorithm.This algorithm can be applied in the author's working university,putting them into Test Paper module of examination management.All those Test Papers are more reasonable and have high partition degree.Meanwhile,this algorithm sharply declines the percentage of genetic algorithm while it enters local optimization.Local searching ability attains further improvement,the examination paper finally cut down detrimental influence for its quality or so.Through a large number of and long time usage,it can be proved that this algorithm take full advantage of time efficiency with high quality.The constitution of the examination paper becomes more scientifically and reasonably in every aspect.