彩印SNS用户亲密度模型的设计与实现
|
摘要
随着互联网技术的快速发展,SNS(Social Network Service)呈现出多样化,渗入到人们生活中游戏、阅读、音乐等领域。然而,这些SNS服务虽然业务形式不同,但是其中的用户关系形式仍然单一,信息过载问题同样存在。这样一来,人们在SNS中所使用的都是简单的人际关系管理,无法进行精细化的关系管理。这导致了SNS对用户而言不过是娱乐性消磨时间的工具,其所应具有的社会交往价值得不到真正的实现。
彩印业务作为一项新兴的电信增值业务,具有移动互联网业务应用的特性。彩印内容是不超过45个字的文本内容,与微博、QQ等签名类似,在互联网上具有良好的传播性。彩印SNS系统就是基于彩印业务搭建的社交服务环境,用户在其中可以评论、复制好友的彩印,也可以向好友推荐自己喜欢的彩印内容。
在了解了SNS所面临的问题和彩印业务的特点之后,通过对彩印内容形式的分析,依据用户行为信息,利用各项分析挖掘技术,提出了一种基于彩印SNS的用户关系量化方法——用户亲密度模型。该模型通过对用户在使用彩印业务的过程中产生的各项行为数据进行分析,并结合用户基础信息,对用户间的亲密度进行计算,形成侧重点不同的各项计算结果,包括基本亲密度、相似亲密度和互动亲密度。
本文分七部分内容：第一章介绍了现在SNS用户关系的研究现状,针对用户关系和信息管理等亟待解决的问题提出了用户亲密度这一解决方案,并介绍了论文结构；第二章对本研究所用到的数据挖掘相关技术进行了简单介绍；第三章分别描述了彩印业务用户数据和彩印源数据；第四章介绍了针对用户源数据和彩印源数据进行分析挖掘的具体方法,搭建本研究所需的数据仓库并加载基本信息；第五章详细描述了用户亲密度模型中基本亲密度、相似亲密度和互动亲密度的计算方法,并介绍了基于这些计算结果形成的综合亲密度的应用场景；第六章对用户亲密度模型进行了检验与测试；第七章是全文的总结和展望。
According to the rapid expansion of Internet technology, a variety of social network services come into people's life in the field of computer games, music, reading and so on. However, those social network services have too much in common, sharing the same problems of dull users relationship and information overloading. In this way, people can not manage the relationship meticulousry, only using the same way. Therefore, social network service only means entertainment for people to kill time, failing to realize it's real value of showing the same association between people in real world.
As a new added-value telecommunication service, Caiyin also has a character which Internet services have in common. The content of a Caiyin is a character string with a length of45. Just like the state words in QQ and twitter, Caiyin will be spread at a high rate of speed in Internet. The SNS of Caiyin is a friendly environment for users of Caiyin service, where they can copy and comment on the Caiyins of their friends, and recommend their favourite Caryin to their friends.
After the analysic for problems in SNS and characteristic of Caiyin service, Users Close Degree Model is required for analyzing users' behavior with a number of data mining tools. The model will provide a good way to tell the differences in relationship between users in Caiyin service, using the calculated results of different kinds of degree including Basis Close Degree, Similarity Close Degree and Interaction Close Degree, based on the data of users'basic information and behavior information.
This paper is divided into seven parts. The first part introduces the current situation of the research about people's relationship in SNS, and recommend the Users Close Degree Model which will solve the problems in users'dull relationship and information overloading. The second part talks about some data mining technologies. The third part tells the data of users and contents in Caiyin service. The fourth part introduces a certain way to construct a data warehouse based on the data of users and contents. The fifth part gives a detail description about the User Close Degree Model which consists of Basis Close Degree, Similarity Close Degree and Interaction Close Degree. And some applications of combination these degrees are also be introduced. The sixth part talk about some ways to examine the model. The last part gives a summary for this paper.
彩印业务作为一项新兴的电信增值业务,具有移动互联网业务应用的特性。彩印内容是不超过45个字的文本内容,与微博、QQ等签名类似,在互联网上具有良好的传播性。彩印SNS系统就是基于彩印业务搭建的社交服务环境,用户在其中可以评论、复制好友的彩印,也可以向好友推荐自己喜欢的彩印内容。
在了解了SNS所面临的问题和彩印业务的特点之后,通过对彩印内容形式的分析,依据用户行为信息,利用各项分析挖掘技术,提出了一种基于彩印SNS的用户关系量化方法——用户亲密度模型。该模型通过对用户在使用彩印业务的过程中产生的各项行为数据进行分析,并结合用户基础信息,对用户间的亲密度进行计算,形成侧重点不同的各项计算结果,包括基本亲密度、相似亲密度和互动亲密度。
本文分七部分内容：第一章介绍了现在SNS用户关系的研究现状,针对用户关系和信息管理等亟待解决的问题提出了用户亲密度这一解决方案,并介绍了论文结构；第二章对本研究所用到的数据挖掘相关技术进行了简单介绍；第三章分别描述了彩印业务用户数据和彩印源数据；第四章介绍了针对用户源数据和彩印源数据进行分析挖掘的具体方法,搭建本研究所需的数据仓库并加载基本信息；第五章详细描述了用户亲密度模型中基本亲密度、相似亲密度和互动亲密度的计算方法,并介绍了基于这些计算结果形成的综合亲密度的应用场景；第六章对用户亲密度模型进行了检验与测试；第七章是全文的总结和展望。
According to the rapid expansion of Internet technology, a variety of social network services come into people's life in the field of computer games, music, reading and so on. However, those social network services have too much in common, sharing the same problems of dull users relationship and information overloading. In this way, people can not manage the relationship meticulousry, only using the same way. Therefore, social network service only means entertainment for people to kill time, failing to realize it's real value of showing the same association between people in real world.
As a new added-value telecommunication service, Caiyin also has a character which Internet services have in common. The content of a Caiyin is a character string with a length of45. Just like the state words in QQ and twitter, Caiyin will be spread at a high rate of speed in Internet. The SNS of Caiyin is a friendly environment for users of Caiyin service, where they can copy and comment on the Caiyins of their friends, and recommend their favourite Caryin to their friends.
After the analysic for problems in SNS and characteristic of Caiyin service, Users Close Degree Model is required for analyzing users' behavior with a number of data mining tools. The model will provide a good way to tell the differences in relationship between users in Caiyin service, using the calculated results of different kinds of degree including Basis Close Degree, Similarity Close Degree and Interaction Close Degree, based on the data of users'basic information and behavior information.
This paper is divided into seven parts. The first part introduces the current situation of the research about people's relationship in SNS, and recommend the Users Close Degree Model which will solve the problems in users'dull relationship and information overloading. The second part talks about some data mining technologies. The third part tells the data of users and contents in Caiyin service. The fourth part introduces a certain way to construct a data warehouse based on the data of users and contents. The fifth part gives a detail description about the User Close Degree Model which consists of Basis Close Degree, Similarity Close Degree and Interaction Close Degree. And some applications of combination these degrees are also be introduced. The sixth part talk about some ways to examine the model. The last part gives a summary for this paper.
引文
[1]中国互联网络信息中心.中国互联网络发展状况统计报告.[2012-1].
[2]艾瑞咨询集团.中国社交化电子商务专题报告2011年.[2012-1].
[3]SENECAL S, NANTEL J. The influence of online product recommendations on consumers'online choices [J].Journal of Retailing,2004,80:159-169.
[4]Ni P, Liao JX, Wang C, Ren KY. Web information recommendation based on user behaviors.2009 WRI World Congress on Computer Science and Information Engineering.2009.3.31-2009.4.2:426-430.
[5]乔秀全,杨春,李晓峰,陈俊亮.基于用户上下文的信任度计算方法的研究与实现[J].计算机学报,2011(12):3-9.
[6]http://tech.sina.com.cn/i/2012-01-18/13286651169.shtml
[7]刘晶晶,王凌娜.基于E-Services的商务社交网络的服务匹配算法[J].东方企业文化,2011(2).:1-3
[8]Thomas H.Comrmen, Charles E.Leiserson.etc, Introduction to Algorithms, Second edition.2006.9
[9]曾春,邢春晓,周立柱.个性化服务技术综述[J].软件学报.Vol 13, No.10, 2002.
[10]高晓琴,蒋朝学,涂瑞.Web使用挖掘研究[J].微计算机信息,2006(21).:1-3
[11]史忠植著.知识发现.清华大学出版社.2002
[12]POPESCU A M,ETZIONI O,Extracting product features and opinions from reviews[C]//Proceedings of the Conference on Human Language Technology and Empirical Methods in Natural Language Processing,Stroudsburg,DA,USA: Association for Computational Ling uistics,2005:339-346
[13]Soo-Min Kim, Eduard Hovy. Determining the Sentiment of Opinions. The 20th Internation Conference on Computational Linguistics[C]. Geneva, Switzerland. 2004.8.29
[14]Lu Y, Zhai C, Sundaresan N, Rated Aspect Summarization of Short Comments, In:Proceedings of the 18th international conference on world wide web, Madrid, Spain:ACM,2009.4.20-24,131-140
[15]Bo Pang, Lillian Lee, ShivakumarVaithyanathan. Thumbsup? Sentiment Classification using Machine Learning Techniques. Proceedings of EMNLP-02,79-86
[16]Yi J,Niblack W, Sentiment mining in Web Fountain [A],In:Proceedings of the 21st International Conference on Data Engineering(ICDE 2005) [C],Washing,DC,U SA:IEEE Computer Society Press,2005,1073-1083
[2]艾瑞咨询集团.中国社交化电子商务专题报告2011年.[2012-1].
[3]SENECAL S, NANTEL J. The influence of online product recommendations on consumers'online choices [J].Journal of Retailing,2004,80:159-169.
[4]Ni P, Liao JX, Wang C, Ren KY. Web information recommendation based on user behaviors.2009 WRI World Congress on Computer Science and Information Engineering.2009.3.31-2009.4.2:426-430.
[5]乔秀全,杨春,李晓峰,陈俊亮.基于用户上下文的信任度计算方法的研究与实现[J].计算机学报,2011(12):3-9.
[6]http://tech.sina.com.cn/i/2012-01-18/13286651169.shtml
[7]刘晶晶,王凌娜.基于E-Services的商务社交网络的服务匹配算法[J].东方企业文化,2011(2).:1-3
[8]Thomas H.Comrmen, Charles E.Leiserson.etc, Introduction to Algorithms, Second edition.2006.9
[9]曾春,邢春晓,周立柱.个性化服务技术综述[J].软件学报.Vol 13, No.10, 2002.
[10]高晓琴,蒋朝学,涂瑞.Web使用挖掘研究[J].微计算机信息,2006(21).:1-3
[11]史忠植著.知识发现.清华大学出版社.2002
[12]POPESCU A M,ETZIONI O,Extracting product features and opinions from reviews[C]//Proceedings of the Conference on Human Language Technology and Empirical Methods in Natural Language Processing,Stroudsburg,DA,USA: Association for Computational Ling uistics,2005:339-346
[13]Soo-Min Kim, Eduard Hovy. Determining the Sentiment of Opinions. The 20th Internation Conference on Computational Linguistics[C]. Geneva, Switzerland. 2004.8.29
[14]Lu Y, Zhai C, Sundaresan N, Rated Aspect Summarization of Short Comments, In:Proceedings of the 18th international conference on world wide web, Madrid, Spain:ACM,2009.4.20-24,131-140
[15]Bo Pang, Lillian Lee, ShivakumarVaithyanathan. Thumbsup? Sentiment Classification using Machine Learning Techniques. Proceedings of EMNLP-02,79-86
[16]Yi J,Niblack W, Sentiment mining in Web Fountain [A],In:Proceedings of the 21st International Conference on Data Engineering(ICDE 2005) [C],Washing,DC,U SA:IEEE Computer Society Press,2005,1073-1083
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |