首页> 外文会议>中国机械工程学会年会暨中国工程院机械与运载工程学部首届年会 >COLLABORATIVE FILTERING RECOMMENDATIONALGORITHM BASED ON LOOK-AHEAD SELECTIVESAMPLING
【24h】

COLLABORATIVE FILTERING RECOMMENDATIONALGORITHM BASED ON LOOK-AHEAD SELECTIVESAMPLING

机译:基于前瞻选择性采样的协同过滤推荐算法

获取原文
获取原文并翻译 | 示例

摘要

Personalized Recommendation System has become anrnimportant research item to prove the suitable product andrnservices for individual. And classification of customersrnbecomes the basis to produce recommendation. In a realisticrnEC system, the magnitudes of customers and products are allrnhuge, so the quality of recommendation decreasesrndramatically. To improve recommending quantity, arncollaborative filtering model was proposed based on lookaheadrnsampling. In n-dimension Euclid space constituted byrnusers, the proposed algorithm reduces the number of samplesrnwhile maintaining the quality of classification, throughrnestimating sample's utility for classifier. At last, experimentsrnwere designed at the basis of MoveLens dataset. Comparedrnwith general collaborative filtering, the proposed algorithmrnhas higher quality of recommendation.
机译:个性化推荐系统已成为一项重要的研究项目,以证明适合个人的产品和服务。并将客户分类成为产生推荐的依据。在一个现实的EC系统中,客户和产品的规模很大,因此推荐的质量急剧下降。为了提高推荐量,提出了基于前瞻性抽样的arncollaborative过滤模型。在由用户组成的n维Euclid空间中,该算法通过估计样本在分类器中的效用,在保持分类质量的同时减少了样本数。最后,在MoveLens数据集的基础上进行了实验设计。与一般协同过滤相比,该算法具有较高的推荐质量。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号