Processing OpenCitations Data
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

Processing OpenCitations Data

This repository processes the OpenCitations data to make it more user-friendly and concise. The primary output is data/citations-doi.tsv.xz, which is a catalog of DOI-to-DOI citations. The file is formated like:

source_doi target_doi
10.1002/14651858.cd002244.pub4 10.1001/archneur.1990.00530120057010
10.1002/14651858.cd002244.pub4 10.1002/14651858.cd002244
10.1002/14651858.cd002244.pub4 10.1002/14651858.cd002244.pub2
10.1002/14651858.cd012199 10.1001/jama.295.6.676
10.1002/14651858.cd012199 10.1001/jama.299.16.1937
10.1002/14651858.cd012199 10.1002/14651858.cd000371.pub6
10.1002/14651858.cd012199 10.1002/14651858.cd009382.pub2

All DOIs are lowercase. Quality control steps were performed on the DOIs to remove clearly incorrect DOIs. However, for best results, we recommend users intersect these DOIs with a catalog of valid DOIs to remove any remaining errant DOIs.


The downloading and processing of the OpenCitations data is accomplished by sequentially running the notebooks in this repository. To update the pipeline to use newer versions of OpenCitations data, one should update the figshare article IDs in


This repository uses conda to manage its environment as specified in environment.yml. Install the environment with:

conda env create --file=environment.yml

Then use source activate opencitations and source deactivate to activate or deactivate the environment. On windows, use activate opencitations and deactivate instead.

In addition, to the conda environment, users will need to install the Disk ARchive (dar) utility to their system.