A guide to putting your research data online using the Datasets API.
Datasets appear on data.mendeley.com and contain one or more research data files as well as metadata describing the files. The files are uploaded via the file contents endpoint. A dataset is initially a draft and can be made publicly available where it will get assigned a DOI. More…
In order to attach a file to a dataset, you first need to upload it to the file contents endpoint. You will receive a file content id in the response, which you can then use to attach this file to the desired object. More…
Dataset objects represent a collection of research data files, with metadata explaining the research data. The files themselves are uploaded via the
/file_contents endpoint of the API, and are then linked to the dataset.
All datasets begin as a draft where they can be edited and shared privately. Once a dataset is made public, it gets associated a Digital Object Identifier (DOI) and will be visible to all users on Mendeley Data.
Datasets are versioned, and every time a dataset is made public a new version is created. Public versions can not be edited, only the draft can be edited.
All endpoints which return a list of results will return only IDs by default. Depending on your implementation, you may want to receive more fields. This is done by a query parameter.
fields query parameter allows you to specify which fields you wish to receive for dataset(s). For example, a request to
/datasets?fields=title,description will result in an array of public datasets including the fields
description. For nested objects such as contributors, you must provide the keys within the object, for example
/datasets?fields=contributors.first_name,contributors.last_name will return the contributors array with
last_name fields within each object.
Whilst it isn't possible for a public dataset to be deleted, it is possible for datasets to be flagged if they don’t contain content we deem to be of a scientific nature appropriate for Mendeley Data.
Flagged datasets do not get returned in results, but making a request for the dataset directly will result in a
451 response. The response will contain a
message specifying the reason why the dataset was flagged.
Dataset files are uploaded via the file_contents endpoint. The
id returned by this endpoint is then provided to the dataset object in order to associate the file contents with the dataset. One or more files may be associated with a dataset.
The file contents endpoint receives file data, and returns an id for the file. This id can then be linked to an object that supports this, such as a dataset.
Data uploaded to the file contents endpoint will need to be linked to another object such as a dataset in order to be used. It is not possible to download data from the file contents endpoint directly, but it is possible to retrieve the download URL and other metadata such as the contents type via the linked object.