links.bulk.csv

Using a web scraper, we checked the status of internet links detected in scientific papers.

The data schema is detailed below:
* type - In which part of the manuscript the link was found.
* journal - Title of the journal where the paper was published.
* id - Pubmed's primary identifier for the paper.
* year - When the paper was published.
* link - URL parsed from the manuscript text.
* code - HTTP/FTP status code returned when trying to access the URL. (-1 indicates a timeout.)
* flag.uniqueness - Whether the link appears only once in the data. '0' means it is unique.
* newtest - The protocol used to determine the status code in our revised pipeline. Only listed for links that were reevaluated.
* oldcode - The status recorded for this link prior to the pipeline revision. Only listed for links that were reevaluated.