Research on auto-composing test paper system based on improved genetic algorithm

Comosing test paper is an important part of examination system. Based on the researches on coding policy, fitness faction, genetic operation and control parameter, an improved genetic algorithm is advanced for the auto-composing test paper system. It is an efficient way to overcome the premature convergence and the genetic drifting, and at the same time,to prevent the colony coming into the partial optimal solution considering the colony diversity by scaling the fitness faction and building the adaptive crossover probability and mutation probability. Experiment shows that the improved genetic algorithm cold compose test paper more efficiency.