Table of Contents
This is an article for a legacy feature of the Data & Insights platform. For more information on how to to create custom metadata on the platform, please visit this article on Metadata Templates.
Data & Insights contains a number of default metadata fields, but many organizations have their own metadata standards, so we have a tool where you can create custom metadata to meet those organizational needs.
Custom metadata is broken down into fieldsets and fields. A fieldset is a way to logically and visually group one or more custom fields together.
Fields can be configured in multiple ways. They can be required, which will be enforced when publishers create new datasets and update existing metadata. They can also be private, meaning that the values entered will not be visible to the public.
Finally, you can choose to pre-define the options for your fields if you wish to have values chosen from a list rather than entered as free-form text.
Creating Custom Metadata
In order to create Custom Metadata, an Administrator or Designer will navigate to the Metadata page within the Administration Panel on your Data & Insights site.
Once on the Metadata page, you will first need to create a custom Fieldset, which can contain one or more specific fields.
Once the Fieldset is created, you will need to create a Field. The field will be how the metadata is described when viewing the dataset. After the field has been created you will see a few more options.
- Make Required - Makes the field required, datasets will not be published until this field is filled out.
- Make Private - Makes the field inaccessible to public users.
- Up/Down - Change the ordering of the field within the fieldset
- Remove - Deletes the field completely.
To set this list simply click the "+" mark under Allowed Values and enter the value. You can add additional values by using the "+" symbol which appears under the current value. You can remove the value by hitting the "x" next to the value.
Deleting Custom Metadata Fields
At any point, you can delete custom metadata field and fieldsets. Removing an entire fieldset will also remove all fields. Please note that there is no way to undo these changes.
It is important to note that removing fields or fieldsets from the metadata page will not remove the metadata from the datasets themselves. While the fields will not be available for new datasets, previous datasets containing these fields will have to do be modified individually or in bulk through the Metadata API.
Enhanced Custom Metadata Options
For some of our users, required asset metadata may include entering codes or other values that are hard for ordinary users to remember or interpret.
For example, one domain might use a required metadata field called "Department Code" which includes values like "1001" to represent the actual departments. To accommodate the needs of data publishers, we offer an optional metadata customization feature that allows site administrators to configure labels for otherwise obscure metadata values on a field-by-field basis.
For example, when the domain administrator configures the custom metadata field "Department Code," when adding the list of allowable values, such as "1001", they can also add meaningful labels for them, like "Department of Information Technology". This means that data publishers who are filling out asset metadata can choose from a list of meaningful labels, like "Department of Information Technology," rather than having to remember or lookup which code corresponds to that department.
Once published, the metadata value which appears in the data.json output, on dataset primer pages, and in all other platform UI surfaces will be the required value ("1001"). This feature simplifies the metadata configuration step for data publishers while ensuring that the published metadata meets requirements.
These enhanced Custom Fields can be marked as private or made required in the same way as any other custom metadata field on your site.
Defining enhanced custom metadata fields.
Selection of metadata based on field label.