[Dataset] Analysis of ego-networks of two CS-related Twitter accounts

Explanation/Overview:

This is the dataset for the analyses and results on Twitter Ego-Networks of two CS-related accounts (@EuCitSci & @SciStarter). The username have been anonymised.

Purpose:

The purpose of this dataset is to provide the basis to reproduce the results reported in the associated deliverable. As such, it does not represent raw data, but rather files that already include certain analysis steps (like calculated degrees or other SNA-related measures), ready for analysis, visualisation and interpretation with R or any other network visualisation software (e.g., Gephi). The edges represent the follow relation and were retrieved using the Twitter API. All usernames except those of the two ego-accounts were anonymised by assigning a random number to each node. Due to the size of the network, we do not include any .gexf or .gml files in this upload, but rather resort to node and edge lists in the .json format

Relatedness:

The networks are the ego-networks for two related public accounts that are associated with citizen science (@EuCitSci & @SciStarter).

Content:

In this Zenodo entry, two files can be found.

edges.json

Represents the edge list, with the columns:

source
target
wherefrom
account

EuCitSci
23929
friendslist
EuCitSci

Source and target are the necessary columns for the network creation and in this example indicate that EuCitSci follows 23929. wherefrom indicates the origin of this relation in the crawling (i.e., whether it was retrieved using the friends or follower list) and account indicates the ego-account it belongs to. Thus, the edges can also be separated using this account attribute as they represent two distinct networks.

nodes.json

Represents the nodes in the networks. The following data fields are contained:

username
CS
followers_count
friends_count
...

1116
CS
689
514
...

...
favourites_count
listed_count
statuses_count
...

...
2601
18
2141
...

...
degree
in_degree
out_degree
...

...
21
5
16
...

...
reciprocity
account

...
0.48
EuCitSci

Username represents the numerical and anonymised username, CS the community-membership. The different counts (e.g., followers_count) indicate the number of followers the user had at the time of the retrieval by the Twitter API. degree refers to the degree in the network (similarly the in- and out_degree), while reciprocity refers to the number of mutual edges in respect to the total number of edges per node (see here). Account is similar as described above.

Grouping:

The data is grouped according the ego-account it is associated to, as can be read above (i.e., the account attribute).