| 基于借阅时间评分的协同图书推荐模型与应用 |
| |
| 引用本文： | 景民昌,于迎辉.基于借阅时间评分的协同图书推荐模型与应用[J].图书情报工作,2012,56(3):117-120. |
| |
| 作者姓名： | 景民昌 于迎辉 |
| |
| 作者单位： | 1. 中国科学院研究生院;2. 中国石油大学图书馆; |
| |
| 摘 要： | 针对数字图书馆系统中的读者借阅历史数据，提出利用图书借阅时间计算读者兴趣度的评分数据模型。该模型把读者的借阅行为分为4种不同的类型：续借、超期借阅（长期超借与短期超借）、正常借阅、盲目借阅，设置阈值，分别计算其相对借阅时间，得出读者对图书的兴趣度。在该数据模型的基础上，实现一个基于协同过滤技术的图书推荐系统，验证模型的有效性。
|
| 关 键 词： | 协同过滤 数据挖掘 图书推荐 用户兴趣度 |
| 收稿时间： | 2011-06-20 |
| 修稿时间： | 2011-11-13 |
CF Recommending Model Based on Borrowing-time Scores and Its Application |
| |
| Jing Minchang,Yu Yinghui.CF Recommending Model Based on Borrowing-time Scores and Its Application[J].Library and Information Service,2012,56(3):117-120. |
| |
| Authors: | Jing Minchang Yu Yinghui |
| |
| Institution: | 1. Graduate University of Chinese Academy of Science,;2. Library of China University of Petroleum,; |
| |
| Abstract: | This paper proposes a new model for calculating interest level based on borrowing-time using readers’ historical data in digital library system.In this model,the reader’s borrowing book behaviors are divided into four types: renewing,extended borrowing,normal borrowing and blind borrowing,with different relative borrowing-time and thresholds to reflect their interest levels.At last,a book recommendation system is implemented based on collaborative filtering to verify the effectiveness of the model. |
| |
| Keywords: | collaborative filtering data mining book recommendation user interest level |
| 本文献已被 CNKI 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 |
