基于知识点的多支持度挖掘算法  

A MULTI-SUPPORT MINING ALGORITHM BASED ON KNOWLEDGE POINTS

作　　者：杨鹤标[1] 刘桂兰[1] 

机构地区：[1]江苏大学计算机科学与通信工程学院,江苏镇江212013

出　　处：《计算机应用与软件》2014年第7期169-172,共4页Computer Applications and Software

基　　金：国家自然科学基金项目(61005017;61202110)

摘　　要：在知识点关联分析方法中,采用单一支持度阈值挖掘频繁知识点集,存在挖掘效率不高的问题。籍此,提出基于知识点的多支持度挖掘算法。算法的思想:针对网络学习平台特有的背景,引入知识点兴趣度和知识点出错频度两个度量因子,用以定量分析学习过程和测试诊断过程,客观地反映用户的学习情况;然后对两个度量因子的相关度进行计算,发现学习过程与测试诊断过程间的相关性;最后,结合多支持度策略,计算出基于知识点度量背景的多支持度,采用改进的多支持度关联规则挖掘进行频繁知识点集的挖掘。实验表明,改进算法在客观的支持度设定基础上,能有效地挖掘出频繁知识点集。In association analysis method in regard to knowledge points, to use the threshold with single support to mine the frequent knowledge points set has the problem of low mining efficiency. Therefore, in this paper we present a knowledge points-based mining algorithm with multi-support. The concept of the algorithm is that, first in consideration of the particular background of web-based teaching platform, we introduce two measure coefficients of knowledge points ： the interestingness and the error frequentness to analysis the learning process and the diagnosis process quantitatively, thus to objectively reflect the learning situation of users. Secondly, we compute the relevancy of interestingness and error frequentness, and find the relevance between the learning process and the diagnosis process. Finally, we figure out the multi-support degree based on knowledge points t background in combination with muhi-support strategy, then mine the frequent knowledge points sets by using the improved multi-support association rules mining algorithm. Experiments show that the improved algorithm can mine the frequent knowledge points sets effectively through setting objective support degree.

关 键 词：知识点 多支持度 关联规则 度量因子 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]