Predicting Tie Strength with Ego Network Structures

This repository contains 41 anonymized Facebook ego networks used in Stolz and Schlereth "Predicting Tie Strength with Ego Network Structures" in Journal of Interactive Marketing (forthcoming). The Facebook ego network data was extracted in combination with a survey. The survey asked users about who they perceived as their closest friends, and who they had recently interacted with on the platform. Furthermore, the data contains a binary matching based on Facebook attributes of ego and alter to assess the similarity of users. These dyad-level binary attributes are stored as csv file. All 41 ego networks are stored as graphml files, named by the identifier of the user (i.e., ego) they belong to. Additionally, the first author's "seed" network is saved separately (without attributes), intended for replication of the sampled perspective. For anonymization, all Facebook identifiers were replaced by independent numeric identifiers.本数据集仓库包含41个经过匿名化处理的Facebook自我中心网络（ego network），该数据集出自Stolz与Schlereth发表于《Journal of Interactive Marketing》（即将刊出）的论文《Predicting Tie Strength with Ego Network Structures》。该Facebook自我中心网络数据集与调查问卷同步提取生成，调查问卷向受访者询问了其认定的亲密好友名单，以及近期在该平台上有过互动的用户名单。此外，数据集包含基于自我（ego）与同伴（alter）的Facebook用户属性构建的二元匹配标记，用以评估用户间的相似程度。这些二元组层面的二进制属性以CSV格式文件存储。全部41个自我中心网络均以GraphML格式文件存储，文件名采用其所属用户（即自我节点）的标识符命名。此外，第一作者的“种子”网络（seed network）单独保存（不包含属性），用于复现抽样视角下的分析流程。为实现匿名化处理，所有原始Facebook标识符均已替换为独立的数字标识符。