The structure of online social networks mirrors those in the offline world
Dunbar, Arnaboldi, Conti and Passarella use the frequency of bi-directional posts to define relationships in two Facebook datasets and a Twitter dataset, building ego-centric networks for more than 185,000 people, and ask whether these have the layered structure found in offline face-to-face networks, which scale in successively inclusive layers at roughly five, fifteen, fifty and 150. The Facebook datasets are best described by a four-layer structure and the Twitter dataset by a five-layer structure. The absolute sizes of those layers and the mean contact frequencies within each match the offline values closely, and all three datasets reveal an innermost layer at about one and a half alters whose existence had only been hypothesised offline. The authors conclude that online communities have very similar structural characteristics to offline face-to-face networks. This vault cites it as the strongest evidence for the layers, and records the objection that clustering a decaying distribution can produce layer-like groups by itself.
Dunbar, Arnaboldi, Conti and Passarella use the frequency of bi-directional posts to define relationships in two Facebook datasets and a Twitter dataset, building ego-centric networks for more than 185,000 people, and ask whether these have the layered structure found in offline face-to-face networks, which scale in successively inclusive layers at roughly five, fifteen, fifty and 150. The Facebook datasets are best described by a four-layer structure and the Twitter dataset by a five-layer structure. The absolute sizes of those layers and the mean contact frequencies within each match the offline values closely, and all three datasets reveal an innermost layer at about one and a half alters whose existence had only been hypothesised offline. The authors conclude that online communities have very similar structural characteristics to offline face-to-face networks. This vault cites it as the strongest evidence for the layers, and records the objection that clustering a decaying distribution can produce layer-like groups by itself.
Written from the abstract, OpenAlex.
What our sources record
Reproduced as it was written when this source was checked, figures and all.
43, 39-47. Two Facebook datasets and a Twitter dataset, more than 185,000 egos, edges defined by frequency of bi-directional posts; four-layer and five-layer structures found, with absolute layer sizes and mean contact frequencies closely matching offline values, plus an innermost layer at about 1.5 alters. doi:10.1016/j.socnet.2015.04.005
Where we use it
- How Many Friends You Can Actually Keep, Why We Gather
Reliability of this reference
The link below was matched against the publisher's record on title, first author and year, and all three agree.
The source
https://doi.org/10.1016/j.socnet.2015.04.005
DOI: 10.1016/j.socnet.2015.04.005