Supporting Social Recommendations with activity based clustering外文翻译.pptVIP

下载本文档

2
0
约5.95千字
约 12页
2015-09-30 发布于江西
举报
版权申诉

Supporting Social Recommendations with activity based clustering外文翻译.ppt

1、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。。
2、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
3、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。
4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
7、VIP文档为合作方或网友上传，每下载1次，网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档

Supporting Social Recommendations withActivity-Balanced Clustering F. Maxwell Harper, Shilad Sen, Dan Frankowski GroupLens Research University of Minnesota Appeared in RecSys?2007 Presented by Sharon HSIAO 2007.12.03 Introduction User group features are intended to facilitate information exchange and social interaction Group interfaces effectively support social recommendations MovieLens was originally built to allow members to rate movies and receive algorithmically generated movie recommendations, they introduced “movie groups” into MovieLens (), a new paradigm of user group features by looking at algorithms for automatically forming and describing user groups the methods of “activity-balanced clustering” and automatic “user group summarization” to algorithmically form and describe the movie groups Background Premises Background: generate a fixed number of clusters, with each cluster containing approximately the same number of users “balanced clustering”: distributing data equally among clusters Premise: movie groups are more interesting and useful if users can understand the characteristics that define and differentiate the groups ACTIVITY-BALANCED CLUSTERING based on their movie rating profiles each group contains about the same number of active users (who has logged in during each of the past three months(1,394); 25% of the total number of users ) active users affect visible contributions to MovieLens: 84% of forum posts, 22% of movie ratings all groups contain approximately the same total number of members Clustering Procedure Evaluation MovieLens users who had logged in during the past year (n=18,760) use (1) standard clustering metrics (2) a comparison with standard k-means clustering on several key indicators* Goal: sociability, K-means might lead to one very successful group, and nine very quiet groups * He, J., Tan, A., Tan, C., Sung, S. On Quantitative Evaluation of Clustering Systems. In Information Retrieval and Clustering, Kluwer Academic