摘要
针对跨社交网络用户匹配数据整合不完善、准确率较低的问题,提出基于多源数据整合的跨社交网络用户匹配方法。通过对多源数据相关挖掘获得用户在不同社交媒体网络中账号的对应关系,采用高斯模型处理账号依据对应关系完成数据整合。拟采用用户名、链接地址、姓名和用户Email信息来表示不同网络用户间的属性相似度,将已知匹配用户的相似度向量作为训练向量,同时将身份匹配转化为二分类问题,计算用户匹配度,完成跨社交网络用户匹配。实验结果表明,上述匹配方法获得的准确率、召回率、F1值等各项指标效果较好,具有一定的应用价值。
At present, incomplete data integration, and low accuracy are the main problems in cross-social network user matching. Therefore, a cross-social network user matching method based on multi-source data integration was proposed. The corresponding relationship of user accounts in different social networks was obtained by mining multi-source data. The Gaussian model was used to process user accounts. According to the corresponding relationship, the data integration was completed. User name, link address, name, and user email were used to represent the attribute similarity between different network users. The similarity vector of known users was taken as the training vector. Meanwhile, the identity matching was transformed into a binary classification problem. Finally, the matching degree was calculated to complete the cross-social network user matching. Experimental results show that the accuracy rate, recall rate, F1 value, and other indexes obtained by the proposed method are better, so this method has a certain application value.
作者
胡三宁
李玉祥
HU San-ning;LI Yu-xiang(College of Applied Engineering,Henan University of Science and Technology,Sanmenxia Henan 472000,China;College of Information Engineering,Henan University of Science and Technology,Henan Luoyang 471000,China)
出处
《计算机仿真》
北大核心
2021年第4期352-355,466,共5页
Computer Simulation
基金
河南省自然科学基金(2018HNZ088)。
关键词
多源数据整合
数据挖掘
高斯模型
属性相似度
二分类问题
Multi-source data integration
Data mining
Gaussian model
Attribute similarity
Binary classification
作者简介
胡三宁(1983-),男(汉族),河南洛宁人,硕士,讲师,研究方向:计算机应用;李玉祥(1984-),男(汉族),河南洛阳人,博士研究生,讲师,研究方向:并行计算和推荐系统。
