CoAuthorNet

CoAuthorNet is a Python package for analyzing and visualizing university authorship networks. It facilitates the retrieval of publication data for university staff, builds co-authorship networks, calculates network metrics, and visualizes the largest connected component of the network. CoAuthorNet is ideal for researchers interested in studying collaboration patterns within academic institutions.

Features

• Fetch publication data for university staff from Google Scholar
• Create bipartite author-publication and co-authorship networks
• Calculate network metrics (average degree, clustering coefficient, shortest path length, etc.)
• Visualize the largest connected component of the co-authorship network

Installation

To install CoAuthorNet, use pip:

pip install CoAuthorNet

Requirements

CoAuthorNet requires the following dependencies:

• pandas
• networkx
• numpy
• matplotlib
• tqdm
• scholarly

These are automatically installed with the package.

Usage

Here’s a quick example to get started with CoAuthorNet.

1. Prepare a CSV file with a column named Staff Name, containing the names of staff members. Example CSV (author_school.csv):

   Staff Name,School,Faculty
   Alice Smith,School of Chemistry,Faculty of Science
   Bob Jones,School of Physics,Faculty of Science
   

2. Run the code below to create and analyze the co-authorship network:

    import CoAuthorNet as yn
    import pandas as pd
   
    staff_data = pd.read_csv("./author_school.csv")
    staff_publications_data = yn.fetch_publications(staff_data, affiliation = 'insert_affiliation')
   
   
    G, author_dict, paper_dict = yn.create_bipartite_network(staff_publications_data)
    author_network = yn.create_authorship_network(G, author_dict)
   
   
    author_network_gc = yn.get_largest_component(author_network)
    yn.save_graph(author_network_gc, "largest_component.graphml")
    metrics = yn.calculate_metrics(author_network_gc)
    print(metrics)
   

Function Reference

• fetch_publications(staff_data, output_csv='staff_publications_data.csv', affiliation)

  • Fetches publication data for each staff member and saves it to a CSV file. Affiliation corresponds to the university or orgnization the individual is connected to. This will allow for more accurate results incase of duplicate names. If unknown, that leave blank.

• create_bipartite_network(df)

  • Creates a bipartite network of authors and publications from the data.

• create_authorship_network(G, author_dict)

  • Converts the bipartite network into a co-authorship network.

• get_largest_component(graph)

  • Extracts the largest connected component of a graph.

• calculate_metrics(G)

  • Calculates various network metrics like average degree, clustering coefficient, etc.

• save_graph(G)

  • Saves largest component as a .graphml file.

• plot_network(G, output_file='largest_component_network.png')

  • Plots and saves the largest connected component of the co-authorship network.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contributing

Contributions are welcome! Feel free to submit a pull request or report issues.