Military Grants

Exploration of publications funded through military grant programs

Minimum requirements

• Python 3.6
• Java (required to run reference matching)

I strongly recommend to use Poetry to manage dependencies. Furthermore, Poetry provides entry points to comfortably run processing pipelines.

Instructions

To get started simply execute

git clone https://github.com/ScholCommLab/military-grants
cd military-grants
poetry install

This will create an isolated virtual environment in your project folder and install all required dependencies.

The processing pipeline is available as follows:

1. poetry run preprocessing
2. poetry run references
3. poetry run articles
4. poetry run metrics
5. poetry run reports

Processing pipeline

Preprocessing

1. Export references from Excel sheets.

• Input: Folder with excel sheets (data/external/input)
• Output: File with all references and grant IDs (data/external/references.csv)

Process references

2. Match references with DOIs

• Input
  • File with all references and grant IDs (data/external/references.csv)
• Output
  • Articles with DOIs that are matched to references (data/processed/articles.csv)
  • Interim: File with one reference per line (data/interim/references.txt)
  • Interim: File containing all results from Crossref (data/interim/reference_matching_results.json)

Process articles

3. Enrich DOIs with Pubmed IDs

• Input
  • Articles (data/processed/articles.csv)
• Output
  • Articles (data/processed/articles.csv)

Collect metrics

4a. Collect altmetrics

• Input
  • Articles (data/processed/articles.csv)
• Output
  • Interim: Response from Altmetric (data/interim/respose_altmetric.csv)*

4b. Collect citations and disciplinary information

• Input
  • Articles (data/processed/articles.csv)
• Output
  • Interim: Response from WoS (data/interim/respose_wos.csv)

4c. Combine results

• Input
  • Results from Altmetric (data/interim/respose_altmetric.csv)
  • Results from WoS (data/interim/respose_wos.csv)
• Output
  • Metrics (data/processed/metrics.csv)

Create results

5. Create results

• Input
  • Articles (data/processed/articles.csv)
  • Metrics (data/processed/metrics.csv)
  • Report template (notebooks/reports/*.ipynb)
• Output
  • Reports (results/*.html)

Acknowledgement

We want to thank Dominika Tkaczyk for all the help. We are also using this project to run the advanced reference matching methods described in [this blog post(https://www.crossref.org/blog/matchmaker-matchmaker-make-me-a-match/)].

This project is based on the cookiecutter data science project template. #cookiecutterdatascience.