If your dataset is a Draft (unpublished), deletions will not be tracked in the dataset’s versions. If either your collection or dataset is a Draft, deletions will be completely removed from Scholars Portal Dataverse.

If your dataset is Published, you cannot delete it, you can only deaccession it. Deaccessioned datasets will no longer be public, but a record of their existence will remain in the repository.

If your collection contains datasets, you cannot delete the collection

Deleting Draft Items

While your collection or dataset is in Draft form, use the following steps to delete items. Deleted items cannot be restored.

Deleting Files from a Draft Dataset

1. Open the Draft dataset which contains the file(s) you want to delete.
2. Check the box beside the name of the file(s) you want to delete.
   
3. Click the Edit Files button and select Delete.
4. Confirm you want to delete the file(s) by clicking on the Delete button on the Delete Files window that appears.
   1. After the screen refreshes, the file(s) you selected for deletion will no longer appear in the dataset’s file list.

Deleting a Draft Dataset

1. Open the Draft dataset you want to delete.
2. Click the Edit Dataset button in the top, right corner.
   
3. Select Delete Dataset from the drop-down menu.
4. Confirm you want to delete the dataset by clicking the Continue button in the Delete Dataset page that appears.
   1. Once the screen refreshes, you will be taken back to the collection’s main page and the dataset will no longer appear.

Deleting a Draft Collection

Note that only empty collections can be deleted. If the collection contains a Draft dataset, refer to the Deleting a Draft Dataset section for instructions on how to remove it.

1. Open the Draft collection you want to delete.
2. Click on the Edit button in the top right corner.
   
3. Select Delete Dataverse from the drop-down menu.
4. Confirm you want to delete the collection by clicking the Continue button on the Delete Dataverse page that appears.
   1. Once the screen refreshes, you’ll be taken to the collection above the one you deleted in the hierarchy.

Deleting Files from a Published Dataset

1. Open the Published dataset that contains the file(s) you want to delete.
2. Check the box beside the name of the file(s) you want to delete.
   
3. Click on the Edit Files button and select the Delete option from the drop-down menu.
4. Confirm you want to delete the file(s) by clicking on the Delete button in the Delete Files page that appears.
   1. Depending on the versions of the dataset, you may receive a warning that the files will only be removed from the current version, but not previous versions of the dataset.
5. Click the Publish Dataset button to republish the dataset and allow the changes to take effect.
   1. This button will only appear after you’ve deleted the file(s) from the dataset.
6. Confirm you want to republish the dataset by clicking the Continue button on the Publish Dataset page that appears.
   1. This page will also display information regarding how the versions will be updated due to the deletion of the file(s).
   2. File deletions will force a Major Release in the versions of the dataset.

Deaccessioning a Published Dataset

Once published, a dataset (or a dataset version) cannot be deleted, but it can be deaccessioned. A tombstone landing page with the basic citation metadata will always be accessible to the public via the persistent URL (handle or DOI) provided in the citation of the dataset.

Deaccessioning a dataset or a version of a dataset is a very serious action that should only occur if there is a legal or valid reason for the dataset to no longer be accessible to the public. Your institution may have policies around deaccessioning. Datasets (or dataset versions) may be deaccessioned for the following reasons:

• Violations of the Borealis Terms of Use
• Copyright violation
• Legal requirements and proven violations
• Research misconduct (e.g., falsification of data)
• Confidentiality

Deaccessioning is available to users with the Admin or Curator role. If you do not see the functions outlined in the steps below within your dataset, you will need to contact your institution for assistance.

Warning: Deaccessioning Cannot be Reversed

Once a version of a dataset has been deaccessioned, you cannot reinstate that version of the dataset. The version’s metadata will remain as a tombstone record, but the files contained within the version will no longer be publicly available to users. A new version of the dataset can be created and published.

Deaccession a Dataset

1. Open the Published dataset you want to deaccessioned.
2. Click on the Edit Dataset button in the top right corner.
   
3. Select Deaccession Dataset from the drop-down menu.
   
4. On the Deaccession Dataset page that appears, check the box beside the version(s) of the dataset you want to deaccession. To fully deaccession the dataset, select all available versions.
5. Select an option from the What is the reason for deaccession? drop-down menu that is most applicable to your situation.
   1. If you select the Other option, add a reason to the text box below the drop-down menu.
6. Enter more information about the reason for deaccessioning your dataset, or other useful information, to the text box in the middle of the window.
   1. 1. Enter a URL for the dataset’s new location in the bottom-most text box, if you’ve moved the dataset to another repository.
7. Click the Deaccession button.
8. Confirm you want to deaccession the dataset by clicking the Yes button on the page that appears.
   1. Once the screen refreshes, the versions you selected will be deaccessioned and will no longer be available publicly.
      

Viewing Deaccessioned Datasets

Once your dataset has been deaccessioned, the main dataset page will display the latest active version (if one exists). A deaccessioned version of your dataset will have a “Deaccessioned” label under the title of the dataset and a red background under the dataset citation.

As the dataset owner, you will be able to view the files that existed in that version of the dataset, but these files are not available to other users unless they have administrative permissions.

Public, non-administrative users of the repository will be able to see a similar page, without the files listed. This “tombstone” page (and its associated persistent identifier) will remain in the repository.