This repository contains several R scripts for creating and visualizing coauthorship networks. It focuses on an egonetwork centered around a single scholar (me in this example). The nodes of the network represent my coauthors (people with whom I have published at least one paper), and the links between two coauthors are proportional to the number of papers they have co-signed (if any).
1_Extract_Network scrapes data from Google Scholar using the R package scholar to build the network. It takes as input the file Coauthors.csv (an example with my coauthors) and returns two files, co.csv and net.csv, representing the nodes and links of the coauthorship network, respectively.
2_Draw_Network.R takes as input co.csv and net.csv produced by the first script and relies on the package networkD3 to produce the file Coauthorship.html, which can be used to interactively visualize the network.
utils contains two functions (used in 1_Extract_Network) to identify and remove near-duplicate publication titles based on character-level similarity and then count how many titles are similar above a given threshold between two sets of publications.
If you need help, find a bug, or want to give me advice or feedback, please contact me!
This repository is mirrored on both GitLab and GitHub. You can access it via the following links:
- GitLab: https://gitlab.com/maximelenormand/coauthorship-network
- GitHub: https://github.com/maximelenormand/coauthorship-network
The repository is archived in Software Heritage: