Abstract:

Aiming at low success rate and efficiency of exam paper generation for small exam database, the paper proposed a prediction calculation based intelligent exam paper generation algorithm without backtracking. Based on prediction calculation, according to the constraints, this algorithm firstly figures out the allocation matrix that meets the requirements of the exam database, and then accesses the exam database. This can improve the success rate of exam paper generation and there is no backtracking when the exam database is accessed. Besides, according some strategy, a batch extraction scheme is used to speed up exam paper generation. This algorithm has a certain universality and high research value for constructing intelligent exam system.

Key words: intelligent exam paper generation;prediction calculation;multiple constraints