摘要
针对传统的移动学习资源自动整合方法含有过多无效数据,存在资源整合所需时间较长、整合后的无效资源过多等问题,提出一种基于片段式移动学习资源自动整合方法。采用不完全资源管理器中的片段式资源特征进行相似度度量,运用近邻传播算法对不完全资源进行聚类,根据聚类结果将同一类资源对象划分到相同簇中,利用同一类对象的属性值对缺失值进行填充。将获取的完整资源集中的各资源进行小波分解,获取小波系数,通过该系数对各个资源的小波方差进行计算,得到小波熵,运用小波域移动学习资源整合算法以小波熵作为资源融合权值,对得到的系数进行整合,经过小波逆转换得到移动学习资源自动整合结果。实验结果证明,所提方法能够缩短整合所需时间,减少整合过程中出现的无效资源,从而提升整合的效果。
Aiming at the problems of excessive invalid data in traditional automatic integration method of mobile learning resources, such as long time for resource integration and excessive invalid resources after integration, this paper proposes an automatic integration method of mobile learning resources based on fragments. The fragmented resource features in incomplete resource managers are used to measure similarity, and the nearest neighbor propagation algorithm is used to cluster incomplete resources. According to the clustering results, the same class of resource objects are divided into the same cluster, and the missing values are filled by the attribute values of the same class of objects. The wavelet coefficients are obtained by decomposing the resources in the complete resource set and calculating the wavelet variance of each resource through the coefficients. The wavelet entropy is obtained. The coefficients are integrated by using the wavelet entropy as the weights of resource fusion in the mobile learning resource integration algorithm in the wavelet domain. The coefficients are reversed by the wavelet transform. In exchange for the result of automatic integration of mobile learning resources. The experimental results show that the proposed method can shorten the integration time, reduce the invalid resources in the integration process, and improve the integration effect.
作者
李景丽
王怀宇
LI Jing-li;WANG Huai-yu(Information Engineering Institute,Baoding University,Baoding Hebei 071000)
出处
《计算机仿真》
北大核心
2019年第11期374-377,共4页
Computer Simulation
关键词
相似度度量
近邻传播算法
小波系数
小波熵
Similarity measure
Nearest neighbor propagation algorithm
Wavelet coefficient
Wavelet entropy
