基于深度学习的移动社交网络用户分类研究  

USER CLASSIFICATION IN MOBILE SOCIAL NETWORK BASED ON DEEP LEARNING

作　　者：王嘉祺 何新磊 汪天一 王新[1,2] Wang Jiaqi;He Xinlei;Wang Tianyi;Wang Xin(School of Computer Science,Fudan University,Shanghai 201203,China;Shanghai Key Laboratory of Intelligent Information Processing,Fudan University,Shanghai 200433,China;Beijing Bytedance Technology Co.,Ltd.,Beijing 100086,China)

机构地区：[1]复旦大学计算机科学技术学院,上海201203 [2]上海市智能信息处理重点实验室(复旦大学),上海200433 [3]北京字节跳动科技有限公司,北京100086

出　　处：《计算机应用与软件》2018年第12期42-48,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(71731004)

摘　　要：随着移动设备和移动服务的高速发展,移动社交网络融入了人们的日常生活。用户分类系统在不同的社交网络中发挥着重要的作用,例如恶意账号检测、高影响力用户发现及会员用户发现。引入深度学习技术来解决用户分类问题,且使用了陌陌的真实数据进行评估。结果表明,对于不同的分类目标,均可取得较好的效果。With the rapid development of mobile devices and mobile services, mobile social networks have become very popular in our daily-life. User classification plays an important role in the different kinds of social networks, such as malicious account detection, discovery of users with high influence and finding VIP users. In the paper, we introduced deep learning to solve the problem of user classification. By using the real data collected from Momo, we evaluated our framework. The result shows that good results can be achieved for different classification targets.

关 键 词：移动社交网络 深度学习 用户分类 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]