基于忠诚度的社交网络用户发现方法

发布时间：2019-01-08 17:49

【摘要】：针对社交网络中提高用户的高黏性问题,提出了一种基于用户忠诚度的用户发现的算法。该算法利用双重RFM模型对用户忠诚度进行计算,挖掘出忠诚度不同分类的用户。首先,通过双重RFM模型动态计算出用户在某一时间段的消费价值与行为价值,得到用户某一时间段的忠诚度;其次,根据用户的忠诚度,确定标度曲线,利用相似度计算找到典型的忠诚用户与不忠诚用户;最后,采用基于模块度的社区发现与独立级联传播模型,发现潜在的忠诚用户与不忠诚用户。在某社交网络的微博数据集上,实现了社会性网络服务(SNS)下用户忠诚度的量化表示,获得了基于用户忠诚度的用户发现结果。实验结果表明,所提算法能够有效挖掘出基于忠诚度的用户分类,可以为社交网站针对用户的个性化推荐及营销等,提供理论支持和实用方法。
[Abstract]:In order to improve the high viscosity of users in social networks, a user discovery algorithm based on user loyalty is proposed. The dual RFM model is used to calculate the loyalty of users, and the users with different categories of loyalty are mined. Firstly, the user's consumption value and behavior value in a certain time period are calculated dynamically by double RFM model, and the loyalty of the user in a certain time period is obtained. Secondly, according to the loyalty of users, the scale curve is determined, and the typical loyal users and disloyal users are found by similarity calculation. Finally, a modular based community discovery and independent cascade propagation model is used to find potential loyal and disloyal users. Based on Weibo data set of a social network, the quantitative representation of user loyalty under social network service (SNS) is realized, and the results of user discovery based on user loyalty are obtained. Experimental results show that the proposed algorithm can effectively mine user classification based on loyalty, and can provide theoretical support and practical methods for personalized recommendation and marketing of social network sites.
【作者单位】：中国石油大学(北京)石油数据挖掘北京市重点实验室;中国石油大学(北京)地球物理与信息工程学院;北京联合大学商务学院;石大兆信数字身份管理与物联网技术研究院;
【基金】：国家863计划项目(2009AA062802) 国家自然科学基金资助项目(60473125,71572015) 中国石油(CNPC)石油科技中青年创新基金资助项目(05E7013) 国家油气重大专项子课题(G5800-08-ZS-WX) 中国石油大学(北京)克拉玛依校区科研启动基金资助项目(RCYJ2016B-03-001)~~
【分类号】：TP301.6;TP393.09

【相似文献】