学习态势在试题智能推荐中的研究

Research on the learning situation in intelligent recommendation of examination questions

在试题智能推荐的研究中定义了知识点Kpu矩阵、知识点的得分分配Skp模型、学习个体对知识点吸收数量等级kpL、学习个体潜在的学习特征向量M SA等有效模型,针对试题推荐中精准率不够高和学习个体学习效果不够好的状况,基于经典DINA模型,改进建立了学习个体的学习态势SA situation模型,完善个体学习态势组成元组,对DINA-学习态势模型进行了评估,利用数据对学习态势准确计算,大大提高试题推荐的精度,并设计试题智能推荐的整体框架,通过对近几年的数据实验,验证得出试题智能推荐精准率、答卷准确率、召回率等变量之间的关系和规律,证明了基于学习态势的推荐方法有助于提高推荐效果,对个体的学习效果有明显提高,对大数据背景下的智能学习有实际的应用价值和科学意义.

In the research of intelligent recommendation of test questions,some effective models are defined,such as Kpu matrix of knowledge points,Skp model of score allocation of knowledge points,kpL of quantity level of knowledge points absorbed by learning individuals,M SA of potential learning feature vector of learning individuals.In view of the situation that the accuracy rate of test questions recommendation is not high enough and the learning effect of learning individuals is not good enough,SA situation model based on learning individuals is established.The model of DINA-learning situation is constructed to calculate the learning situation accurately,which greatly improves the accuracy of test recommendation.The overall framework of intelligent recommendation was designed.Through the data experiments in recent years,the relationships and rules among the variables of intelligent recommendation accuracy,answer accuracy and recall rate were verified,which proved that the model is based on science.The recommendation method of habitual situation is helpful to improve the recommendation effect and the individua s learning effect.It has practical application value and scientific significance for intelligent learning under the background of large data.