| 基于概念泛化的科技文献推荐算法 |
| |
| 引用本文： | 徐勇,司凤山,吴延辉,陈建国,周善英.基于概念泛化的科技文献推荐算法[J].图书情报工作,2012,56(21):101-108. |
| |
| 作者姓名： | 徐勇 司凤山 吴延辉 陈建国 周善英 |
| |
| 作者单位： | 安徽财经大学管理科学与工程学院 蚌埠 233030 |
| |
| 基金项目： | 教育部人文社会科学研究青年基金项目“科技文献推荐系统若干问题研究” |
| |
| 摘 要： | 针对科技文献特征词在语义上的层次特性,提出基于概念泛化的内容过滤推荐算法.采用矢量空间模型作为用户兴趣偏好和科技文献特征的描述模型;在比较科技文献特征与用户兴趣偏好的相似程度时,首先从字符层面比较科技文献特征词与用户兴趣特征词,然后在基于ODP目录结构的用户兴趣偏好概念泛化树上对字符不相同的特征词对进行语义比较,并修正特征词权重,以避免遗漏"字符不同,但语义相似"的关键词对.理论分析和实验结果表明,该算法能够更加全面、准确地推荐科技文献对象.
|
| 关 键 词： | 科技文献 推荐 ODP目录 概念泛化 |
| 收稿时间： | 2012-06-12 |
Concept Generalization-based Recommendation Algorithm of Academic Papers |
| |
| Xu Yong,Si Fengshan,Wu Yanhui,Chen Jianguo,Zhou Shanying.Concept Generalization-based Recommendation Algorithm of Academic Papers[J].Library and Information Service,2012,56(21):101-108. |
| |
| Authors: | Xu Yong Si Fengshan Wu Yanhui Chen Jianguo Zhou Shanying |
| |
| Institution: | School of Management Science & Engineering, Anhui University of Finance & Economics, Bengbu 233030 |
| |
| Abstract: | Aim to semantic hierarchy characteristics of key words of academic papers, content filtering algorithm based on conceptual generalization is proposed. Vector space model is used to represent user interest profiles and academic paper features. To compute the similarity between academic paper features and user interest profiles, at first, academic papers are compared with user interest profiles in character view; then academic paper features are compared with concept generalization tree of user profiles based on the ODP in semantic view, and the weight of corresponding key word is modified, which could avoid missing key word groups which "characters are different, but semantic are similar". Theoretical analyses and experimental results show that the algorithm can more comprehensively, accurately recommend academic papers. |
| |
| Keywords: | academic paper recommendation ODP catalog concept generalization |
| 本文献已被 万方数据 等数据库收录！ |
| 点击此处可从《图书情报工作》浏览原始摘要信息 |
| 点击此处可从《图书情报工作》下载免费的PDF全文 |
|
