-
Notifications
You must be signed in to change notification settings - Fork 1
Add a record step by step
Login and then start a new upload by clicking on the `+` sign next to your name or if you are in `My dashboard` page on the `New upload` button.
Add ACDG community
As detailed in the Communities page, all records should be submitted to the ACDG community for review. Clicking on the select community
button will show you all available communities.
Select only ACDG
, if you want the record to be also part of a different community please notify us, as currently only one community can be selected from the web interface.
This shows the community you selected, and the button to change/remove it, if you selected the wrong one.
If you have a file related to the record you want to upload, you can do it at the top of the form, if not click the metadata-only
option.
As our storage is currently limited, please limit files to what is strictly essential and valuable.
Examples are:
- files containing additional information not available via official records
- reviews of the resource contributed by the community
- notebooks and other code related to a dataset
Files that are easily available via a link, should be listed instead in the related information section at the bottom of the form.
Before uploading any big ( > 0.5 GB ) file please let the admin know so we can make sure there is enough storage available
Do not upload any data, we only add metadata to the records, never the actual data!
Also keep in mind that once you publish a record you need to create a new version to upload/remove files (see below Creating new version)
DOI
As we are always listing existing data you should always select Yes
.
If the data has never been published use a fake doi, as long as you start the doi with the prefix: 10.1234567/ it will be detected as a fake by the tool and not displayed in citations etc.
If the DOI is a pre-existing (i.e., if you selected “yes”) it can always be updated even for a published record.
If you accidentally selected No
and published the record, you will notice a new DOI has been created. This DOI is not a valid Datacite DOI, however it is still used in the DB as the main identifier for the record. Hence, it is not possible to edit it anymore, as the tool prevents substantial changes to potentially authoritative records. Let the administrators know, they can edit it directly via the DB.
Resource Type
This indicates what kind of resource is described by the record. This is one the field used to query records.
Click on the side arrow to display them and scroll up and down to select one. The main categories are:
- Dataset
- Software
- Web service
For each one of them there are a few sub-categories like `Dataset-model output`
Try to choose the most relevant one, or `Other` as the last resource.
For a full list and some explanation of the terms see our [Vocabularies Schema](https://docs.google.com/spreadsheets/d/1EwskocPDq_UI44vTBw12_i5zqKTszCvzdJFcgAjN7mw/edit#gid=415843847).
Title
The resource title, use the original/official title as much as possible, as this will be used to generate citations automatically. Clearly this is not as important if the resource is not published.
If an acronym or alternative title is commonly used to refer to the resource use Add titles
. Please note you need to select a title type, usually alternative
is fine.
Publication date
As for the title this is used to generate citations so it needs to be accurate (at least the year) if the record is published.
If the record is not published, for example for a data portal use creation date if known or simply leave the current date which is used as default.
For any date format is YYYY-MM-DD
where both month and day can be omitted. If omitted 1st month and 1st day are assumed i.e 1980 is interpreted as 1980-01-01.
Creators
Clicking on the add creator
will open the Add creator
form
All names added here will be considered authors, they can be both a person and/or an institution. At least one name needs to be provided.
You can try to see if a name is already registered by starting typing it, as shown above.
For name identifiers use the full url, so the scheme can be recognised. For example:
`https://orcid.org/0000-0002-1234-5678` instead of `0000-0002-1234-5678`.
Role is usually not necessary as all creators are automatically authors. However, if the creator is an organisation the `rights-holder` role might be appropriate.
Also, if there are several authors use the `ContactPerson` role for the main contact, if known.
As creators are used to build the citation add only the official authors in the correct order (creators rows can be swapped around).
For other contributors use the contributor entry further down the form.
Description
This is the main field where the resource can be described. Tips:
- Start with a short summary of what the data is.
- When the records are listed in the main page the first couple of paragraphs of the description are shown together with the title and authors.
- Try to include a list of variables, the geospatial and temporal boundaries, resolution and frequency whenever applicable.
- Add a reference to the related information you are going to link or files that might be uploaded, so a user understands their relationship to the record.
Generally make sure that by reading this a user can get a good idea of what the resource is without having to download files. It is not uncommon for official metadata records to be quite lacking, so if you start from an existing description try to enhance it if possible.
Additional descriptions
By default we add two additional descriptions
as a reminder to include any available location of the data and relevant access information. If this is not applicable to your record feel free to remove them.
The location
is the path for the data on a local server, for example if the data is at NCI, you can list the url to access the project, and the path. If data is available online the url.
Use the access
description when there is a preferred citation, acknowledgement statements, or other special instructions in regard to the data access.
For example, as shown on the left, often data providers request data users to also cite a paper related to the data:
Other types of additional description are also available, use them wherever you want to make some information more visible.
License
By default the tool will list a Creative Commons Attribution 4.0 International
license. If the record has a different license you can remove this and click on add standard
to browse ready to use licenses or add custom
if the record has a custom license or one that is not included in the list.
The custom license form includes a title
, a description
for the license text, and a link
if one is available.
Contributors
As for creators but this time the role is compulsory, for an explanation of roles meaning see the Datacite Schema.
Dates
There are different kind of dates available, for a complete list refer to the relevant tab in the Datacite schema.
Format for dates is: DATE
or DATE/DATE
where DATE is YYYY
or YYYY-MM
or YYYY-MM-DD
We recommend for all datasets to add from_date
and to_date
whenever possible. Above it is shown an example on how to add them. If a dataset covers two separate temporal ranges, use their union to determine from/to dates.
NB dates is not a required field, you can remove the date box appearing by default if there is no date you want to add.
Version
This should indicate the version of the object you are describing in the record, not of the record itself.
While version
is an optional field you should always add where available as it will be used to form the citation field when present.
Publisher
This is a required field as it is used to create the citation field, by default it will have a Not available
value, which is ok for records that do not have a DOI. If adding a properly published record put here the organisation that minted the DOI, e.g. Zenodo, in a form which is suitable for a citation.
Keywords are important as they are used to query the records, so you should aim to add as many as possible but also to make sure they are descriptive of the record, if none of the available tags are suitable they're better left out.
We created 5 custom fields to label our records:
- Geospatial resolution
- Frequency
- Region
- Variable realm
- Files format
Each custom field has its own controlled vocabulary.
The available terms can be selected from a dropdown menu or you can start typing in the box to subset them, as shown below.
Subjects
As shown on the left subjects are divided in theme each with its own controlled vocabulary (link in keywords section). Currently we only have one theme a subset of the ANZSRC Field Of Research 2020 codes
You can select a theme to limit the terms choices, or choose to select from all the themes available. Once you selected a theme, you can choose from its terms by starting typing in the search box, as shown on the left.
- While it is possible to type a new term,
- please only choose from available ones!
If you want to add a new term, the correct procedure is to open an issue, so this can be discussed and eventually the term added to the most appropriate vocabulary.
Related works is an optional but highly recommended field. You should add here any link to related resources. These could be a code repository, a dataset landing page, direct access to original data, documentation relating to the resource etc.
* `identifier` is the actual link; * `scheme
` is the link type (i.e., doi, url);
-
resource_type
indicates the kind of the linked resource, it uses and extended version of theresource_type
vocabulary used for the record itself. See the original Datacite metadata schema v4.4, pg 48 for a full list of terms.
It is really important to avoid confusion to clarify the relationship with the described resource as clearly as possible. We use theRelation
terms defined by Datacite plus a generic termIsRelatedTo
to use if nothing else is suitable. The table below also lists the appropriate usage for common relationships.
The terms should be interpreted with the described resource as the subject, so for example usingIsCitedBy
and linking a published paper implies that the paper cites the described resource.
Resource | Suggested relationship |
---|---|
the original metadata record, i.e. a geonetwork record for a dataset published by NCI |
IsIdenticalTo or IsVariantFormOf depending on how different from original |
the data portal record or repository for the described data/code | IsMetadataFor |
documentation or training material | IsDocumentedBy |
a review of the resource | IsReviewedBy |
a dataset/code used as base for this resource | IsDerivedFrom |
a dataset/code derived from this resource | IsSourceOf |
a dataset/code required to use this data/code | Requires |
a dataset/code that is superceeded by the resource | Obsoletes |
a paper or web page describing the resource | IsDescribedBy |
At a first look this and the next two fields might appear as required<br>
(i.e., with a red `*`), but they are optional fields. The `*` are<br>
indicating that what information is required if you add a related<br>
work record. Just close the tab if you don't want to add one.
Alternate identifiers
Alternate identifiers are also optional, but should be used for other identifiers of the resource described in the record. This could be useful for resources that have been published with an identifier different from a doi, i.e., with a handle.
Another possible use is for unpublished resources if you want to highlight their more authoritative source.
Visibility
By default visibility is public and there is no embargo on the records. As we can use communities to impose a kind of embargo until the record is ready to be shared. There shouldn't be any need to change any of these settings.
Saving
The fields that are required for a record to be valid are highlighted by a red *
.
You can still save a draft
even if they are not filled, however the tool will print out
a warning and it will highlight in red all the fields that still need to be filled for the record to be valid.
Until you filled them you won't be able to publish the record.
You can preview
the record page before going ahead and publish, NB some features might not be fully rendered in the preview.
Publish
When you are ready, ensuring you have clicked Save draft
, simply click on the Publish
button shown above, you will be asked to confirm. As once a record is published there are some fields that cannot anymore be changed without producing a new version, it is important to make sure the following fields are correct:
- DOI - double check you selected the DOI existing option and that DOI is correct!
- Files - if you have files you want to add do it now, once you publish you cannot add/remove files without creating a new version!
- Title - this can be changed without consequences but it might create confusion to users.
© ACDGuide 2022