covid19_preprints dataset

<strong>This dataset is generated by the code in the Github repository </strong><strong>covid19_preprints. </strong> From March 2023, the dataset is archived separately here due to file size. This also constitutes a citable version of the dataset. <br> The repository contains code used to extract details of preprints related to COVID-19 and visualize their distribution over time. Work by Nicholas Fraser and Bianca Kramer. <br> Preprint data is updated monthly by collecting preprints added or updated since the previous sample date. Details of the current release can be found in <code>data/metadata.json</code>, where <code>release_date</code> refers to the date on which the dataset was released, <code>sample_date</code> the date on which data was collected, and <code>posted_date</code> the cut-off point for preprints to be included in the accompanying visualizations based on their posting date. <br> A citable version of this repository is also available on figshare, here: https://doi.org/10.6084/m9.figshare.12033672. <br> Note that this dataset is not exhaustive, but aims to collate information from some of the main sources of preprint metadata. <br> The process for collecting preprint metadata is documented fully here. In general terms, preprint metadata are harvested from four main sources: Crossref (using the rcrossref package). All records with the <code>type</code> field defined as <code>posted-content</code> and <code>subtype</code> field defined as <code>preprint</code>are harvested, as well as records from SSRN (where the <code>type</code> field is instead defined as <code>journal-article</code>). Preprint records are then matched to known preprint repositories based on <code>institution</code>, <code>publisher</code> and <code>group-title</code> metadata fields. DataCite (using the rdatacite package). All records with the <code>resourceType</code> or <code>resourceTypeGeneral</code> field defined as <code>Preprint</code> are harvested. Preprint records are matched to known preprint repositories based on <code>client</code> fields.<br> Note: as of 2022, arXiv also uses DataCite DOIs, with all backfiles receiving DataCite DOIs over the course of 2022. To prevent including this arXiv backlog, DataCite harvesting is currently limited to ResearchGate, Zenodo and Figshare. arXiv (using the aRxiv package). Records are harvested by searching directly for COVID-19 related keywords in titles or abstracts using the built-in search functionality of the arXiv API. RePEc (using the oai package)). All record types are initally harvested, and subsequently filtered for those with the <code>Type</code> field defined as <code>preprint</code>. NB RePEc is not a preprint server in itself but an indexing server for multiple individual archives.<br> Note: For certain time periods (e.g. between 2021-07-15 and 2021-09-08 and after 2021-12-30) no new preprints got retrieved from RePEc. For all sources, preprints are classified as being related to COVID-19 on the basis of keyword matches in their titles or abstracts (where available). The search string is defined as: <code>coronavirus OR covid-19 OR sars-cov OR ncov-2019 OR 2019-ncov OR hcov-19 OR sars-2</code>. <br> In some cases, multiple preprint metadata records are registered for a single preprint (e.g. ChemRxiv registers a new Crossref record for each new version of a preprint). In these cases, only the earliest posted version is included in this dataset. Depuplication is done based on versioning information in dois as well as on case-insensitive title-matching. Additionally, some preprints are deposited to multiple preprint repositories - in these cases all preprint records are included.