摘要
由于英语学习资源与正常文本或图像的数据属性特征存在一定差异,故无法为用户推荐符合需求的英语学习资源,因此,提出基于XML的英语学习资源协同过滤推荐方法。通过构建用户评分矩阵,设定英语学习资源为词频向量,利用改进的余弦相似性,度量邻域用户之间的相似性,根据解得的作品属性隶属函数,推导出属性特征隶属度矩阵,分别计算XML文档的内容相似度与结构相似度,经加权融合求解作品的综合相似度,实现英语学习资源的个性化推荐。实验结果表明,所提方法推荐结果较为精准,具有有效性与可行性。
Since the data attribute characteristics of English learning resources are different from those of normal text or image,it is impossible to recommend English learning resources that meet the needs of users.Therefore,a collaborative filtering recommendation method of English learning resources based on XML is proposed.By constructing the user score matrix,setting the English learning resources as the word frequency vector,and using the improved cosine similarity to measure the similarity between neighboring users,the attribute feature membership matrix is derived according to the solved work attribute membership function,the content similarity and structural similarity of XML documents are calculated respectively,and the comprehensive similarity of works is solved by weighted fusion,realizing the personalized recommendation of English learning resources.The experiment results show that the recommended results of the proposed method are more accurate and have superior effectiveness and feasibility.
作者
孙萍
SUN Ping(Department of Public Basic Courses,Architecture Labor University of Shaanxi Province,Xi’an 710100,China)
出处
《信息技术》
2023年第10期118-122,共5页
Information Technology
关键词
XML
协同过滤
个性化推荐
内容相似度
结构相似度
XML
collaborative filtering
personalized recommendation
content similarity
structural similarity