[论文解读] Do the Young Live in a "Smaller World" Than the Old? Age-Specific Degrees of Separation in a Large-Scale Mobile Communication Network.

Yuxiao Dong, Omar Ližardo+1 more|arXiv (Cornell University)|Jun 24, 2016

Human Mobility and Location-Based Analysis被引用 5

一句话总结

本研究通过分析数十亿条手机通信记录，探讨大规模社交网络中不同年龄群体的平均分离度，发现年轻人经历的‘最小世界’效应最显著（平均分离度为9.52），而年长成年人则经历最大的社会距离。研究结果表明，网络连通性存在系统性的年龄异质性，其成因在于亲属关系纽带和整体互动频率的差异，而性别差异不显著。

ABSTRACT

That any two persons are separated by a relatively small number of intermediary contacts -- the small-world phenomenon -- is a surprising but well established regularity in human social networks. To date, network science has ignored the question of whether the small world phenomenon manifests itself in similar ways across dyadic classes defined by individual traits, such as age or sex. To address this gap in the literature, we explore the phenomenon of age-specific small worlds by employing a mobile phone network built from billions of communication events approximating interaction patterns at a societal scale. We observe the average distance between any pair of users is 9.52, corresponding to nine-and-a-half degrees of separation. More importantly, we show that there is a systematic relationship between age and the average distance connecting that person to others, with some age groups falling below this average quantity while others falling above. Young people live in the smallest world, being separated from other young people and their parent's generation via a comparatively small number of intermediaries. Older people live in the least small world, being separated from their same age peers and their younger counterparts by a relatively large number of intermediaries. Middle age-people fall in between, being sociometrically close to both younger and older generations. However, there exists no significant difference of this age-effect on small world size between men and women. In all these results demonstrate that age-group heterogeneity of the small world can be traced to well-known social mechanisms affecting the way that age interacts with overall volume of connectivity and the relative prevalence of kin ties and non-kin ties, and may have important implications for our understanding of information cascades, diffusion phenomena, and the localized spread of fads and fashions.

研究动机与目标

• 调查小世界现象是否在人类社交网络中随年龄群体系统性变化。
• 检验年龄如何影响大规模通信网络中个体之间的平均路径长度（分离度）。
• 确定性别是否调节年龄相关的网络连通性差异。
• 识别支撑年龄相关网络小世界特性差异的社会机制，如亲属关系纽带和互动总量。

提出的方法

• 本研究使用包含数十亿条匿名通话记录的大规模移动通信网络，以在社会层面建模人类互动模式。
• 构建一个加权无向网络，其中节点代表用户，边代表通信事件。
• 计算所有用户对之间的平均最短路径长度（分离度），并按年龄组聚合。
• 分析特定年龄组的网络距离，以比较不同年龄群体的平均路径长度。
• 通过分别比较男性和女性的结果，控制性别变量的影响。
• 采用统计建模方法评估年龄和性别对网络连通性的显著性影响。

实验结果

研究问题

• RQ1 不同年龄群体在大规模社交网络中是否经历不同的平均分离度？
• RQ2 小世界特性在年轻群体还是年长群体中更为显著？
• RQ3 亲属关系纽带和整体通信量如何调节年龄相关的网络连通性差异？
• RQ4 在特定年龄群体的分离度上是否存在显著的性别差异？

主要发现

• 所有用户整体的平均分离度为9.52，表明在社会层面存在小世界结构。
• 年轻人经历最小的世界，其与同龄人及父母一代的连接路径长度低于整体网络平均值。
• 年长成年人所处的社会网络最小世界效应最弱，其与同龄人及年轻一代的路径长度相对较大。
• 中年人表现出中等连通性，与年轻一代和年长一代的社会距离均较近。
• 男性与女性在年龄相关的最小世界效应上无显著差异。
• 观察到的网络小世界特性在年龄上的异质性，可归因于整体连通性水平的差异以及亲属与非亲属关系纽带的相对分布差异。