|
|||||
|
|
| 基于改进的双极Slope One算法的农业信息推荐模型研究 | |
| 引用本文： | 张启宇,宋瑶,陈英义,封文杰.基于改进的双极Slope One算法的农业信息推荐模型研究[J].安徽农业科学,2014(29):10394-10396. |
| 作者姓名： | 张启宇 宋瑶 陈英义 封文杰 |
| 作者单位： | 1. 中国农业大学信息与电气工程学院,北京100083;中国农业大学烟台研究院,山东烟台264670 2. 中国农业大学信息与电气工程学院,北京,100083 3. 山东省农业科学院科技信息研究所,山东济南,250100 |
| 摘 要： | 为了在海量信息中为用户提供有价值的信息,个性化在线推荐系统是农业信息化综合服务平台的重要组成部分。Slope One算法因简单高效被许多在线推荐系统使用。对Slope One算法进行研究,并结合农业信息化综合服务平台的特点,提出了农业信息推荐模型。该模型按资源类别把用户评分矩阵划分为评分子矩阵,在此基础上采用杰卡德统一算子距离相似度算法建立用户类别近邻用户,采用改进的双极Slope One算法和Item user average算法相结合的方式对未访问资源进行评分预测,有效降低计算量,并提高了评分预测精度,既适合稠密数据集,又适合稀疏数据集。 |
| 关 键 词： | 杰卡德统一算子距离相似度算法 双极Slope one算法 Item user average算法 农业信息 个性化推荐 |
The Bipolar Slope One Algorithm Based on Improved Agricultural Information Recommendation Model Research |
|
| ZHANG Qi-yu,SONG Yao,CHEN Ying-yi,FENG Wen-jie.The Bipolar Slope One Algorithm Based on Improved Agricultural Information Recommendation Model Research[J].Journal of Anhui Agricultural Sciences,2014(29):10394-10396. | |
| Authors: | ZHANG Qi-yu SONG Yao CHEN Ying-yi FENG Wen-jie |
| Institution: | ZHANG Qi-yu, SONG Yao, CHEN Ying-yi, FENG Wen-jie( 1. College of Information and Electrical Engineering, China Agri- cultural University, Beijing 100083 ; 2. Yantai Academy, China Agriculture University, Yantai, Shandong 264670 ; 3. Institute of Information Technology, Shandong Academy of Agricuhural Sciences, Jinan, Shandong 250100) |
| Abstract: | In order to provide valuable information to users in the vast amounts of information,personalized online recommendation system is an important part of a comprehensive agricultural information service platform. Slope One is simple and high efficient algorithm,which is used by many online recommendation system. The Slope One algorithm was studied,and combined with the characteristics of agricultural information service platform,the proposed recommendation model of agricultural information was put forward. The model of the user is divided into sub-matrices scoring matrix by resource category. On the basis of this,a unified operator by using Jaccard similarity algorithm to create a user class neighborhood user. By using improved bipolar Slope One and Item user average,the score of resources not visit was predicted,which not only can reduce computation,but also can improve the prediction accuracy score,both suitable for dense data sets and sparse data sets. |
| Keywords: | Jaccard similarity algorithm uniform distance operator Bipolar Slope One algorithm Item user average algorithm Agricultural information Personalized recommendation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |