中文摘要：

提出了一种属性掌握概率的分类模型,该模型基于Q矩阵,采用对属性掌握概率先估计后分类的方法,从而实现对考生知识状态的识别.在估计阶段提出一种属性掌握概率的估计方法,在分类阶段引进模糊数学的贴近度按择近原则判别法,并通过计算机模拟研究,发现该模型适用于总体的属性掌握概率服从左偏态分布和双峰分布的考生知识状态的识别.

英文摘要：

A classifying model for probability of attributes＇ mastery is proposed. This is a model of cognitive diagnosis based on Q matrix, which can be used to recognize student knowledge structure, whose mastery probabilities of the attributes obey negatively skewed distribution or bimodal distribution.