Abstract: The intelligence algorithm of auto-generating test paper is the key technology of network examination system. Based on analyzing the quality index and the mathematical model, this paper proposes an intelligent algorithm of assembling test paper based on chaos optimization search. An encoding question bank is constructed to minimize the exchanging times of computation, whereas a segmental-composing encoding method based on question types is introduced to satisfy the constraints of test paper and improve the coding efficiency, then a multi-sequence chaotic search method based on mutation operation is introduced to ensure optimal search of ergodicity and randomness. Finally, the computing performance, the coverage of the paper and the matching degree with expected values are demonstrated by computer simulations.

Key words: auto-generating test paper system, Chaos optimization search module, coding method, mutation operation