The Adobe Experience Platform user interface allows you to create and manage Experience Data Model (XDM) schemas in an interactive visual canvas called the Schema Editor. This tutorial covers how to create a schema using the Schema Editor.
For demonstration purposes, the steps in this tutorial involve creating an example schema that describes members of a customer loyalty program. While you can use these steps to create a different schema for your own purposes, it is recommended that you first follow along with creating the example schema to learn the capabilities of the Schema Editor.
This tutorial requires a working understanding of the various aspects of Adobe Experience Platform involved in schema creation. Before beginning this tutorial, please review the documentation for the following concepts:
The Schemas workspace in the Platform UI provides a visualization of the Schema Library, allowing you to view manage the schemas available for your organization. The workspace also includes the Schema Editor, the canvas on which you can compose a schema throughout this tutorial.
After logging into Experience Platform, select Schemas in the left navigation to open the Schemas workspace. The Browse tab displays a list of schemas (a representation of the Schema Library) which you can view and customize. The list includes the name, type, class, and behavior (record or time-series) on which the schema is based, as well as the date and time the schema was last modified.
See the guide on exploring existing XDM resources in the UI for more information.
To begin composing a schema, select Create schema in the top-right corner of the Schemas workspace. A dropdown menu appears, giving you the option to choose between the core classes XDM Individual Profile and XDM ExperienceEvent. If these classes do not suit your purposes, you can also select Browse to choose from other available classes or create a new class.
For the purposes of this tutorial, select XDM Individual Profile.
Since you chose a standard XDM class to base the schema on, the Add field group dialog appears, allowing you to immediately start adding fields to the schema. For now, select Cancel to exit the dialog.
The Schema Editor appears. This is the canvas upon which you will compose your schema. An untitled schema is automatically created in the Structure section of the canvas when you arrive in the editor, along with the standard fields included in all schemas based on that class. The assigned class for the schema is also listed under Class in Composition section.
You can change the class of a schema at any point during the initial composition process before the schema has been saved, but this should be done with extreme caution. Field groups are only compatible with certain classes, and therefore changing the class will reset the canvas and any fields you have added.
Use the fields on the right-hand side of the editor to provide a display name and optional description for the schema. Once a name is entered, the canvas updates to reflect the new name of the schema.
There are several important considerations to make when deciding on a name for your schema:
This tutorial composes a schema to ingest data related to the members of a loyalty program, and therefore the schema is named “Loyalty Members”.
You can now begin to add fields to your schema by adding field groups. A field group is a group of one or more fields that are often used together to describe a particular concept. This tutorial uses field groups to describe the members of the loyalty program and capture key information such as name, birthday, phone number, address, and more.
To add a field group, select Add in the Field groups sub-section.
A new dialog appears, displaying a list of available field groups. Each field group is only intended for use with a specific class, therefore the dialog only lists field groups that are compatible with the class you selected (in this case, the XDM Individual Profile class). If you are using a standard XDM class, the list of field groups will be intelligently sorted based on usage popularity.
Selecting a field group from the list causes it to appear in the right-hand rail. You can select multiple field groups if desired, adding each one to the list in the right rail before confirming. In addition, an icon appears on the right-hand side of the currently selected field group which allows you to preview the structure of the fields it provides.
When previewing a field group, a detailed description of the field group’s schema is provided in the right-hand rail. You can also navigate through the field group’s fields in the provided canvas. As you select different fields, the right rail updates to show details about the field in question. Select Back when you are finished previewing to return to the field group selection dialog.
For this tutorial, select the Demographic Details field group, then select Add field group.
The schema canvas reappears. The Field groups section now lists “Demographic Details” and the Structure section includes the fields contributed by the field group. You can select the field group’s name under the Field groups section to highlight the specific fields it provides within the canvas.
This field group contributes several fields under the top-level name
person with the data type “Person”. This group of fields describes information about an individual, including name, birth date, and gender.
Remember that fields may use scalar types (such as string, integer, array, or date), as well as any data type (a group of fields representing a common concept) defined within the Schema Registry.
Notice that the
name field has a data type of “Person name”, meaning it too describes a common concept and contains name-related sub-fields such as first name, last name, courtesy title, and suffix.
Select the different fields within the canvas to reveal any additional fields they contribute to the schema structure.
You can now repeat the same steps to add another field group. When you view the Add field group dialog this time, notice that the “Demographic Details” field group has been greyed out and the checkbox next to it cannot be selected. This prevents you from accidentally duplicating field groups that you have already included in the current schema.
For this tutorial, select the “Personal Contact Details” field group from the dialog, then select Add field group to add it to the schema.
Once added, the canvas reappears. “Personal Contact Details” is now listed under Field groups in the Composition section, and fields for home address, mobile phone, and more have been added under Structure.
Similar to the
name field, the fields you just added represent multi-field concepts. For example,
homeAddress has a data type of “Postal address” and
mobilePhone has a data type of “Phone number”. You can select each of these fields to expand them and see the additional fields included in the data type.
The “Loyalty Members” schema is meant to capture data related to the members of a loyalty program, so it will require some specific loyalty-related fields.
There is a standard Loyalty Details field group that you can add to the schema to capture common fields related to a loyalty program. While you are strongly encouraged to use standard field groups to represent concepts captured by your schemas, the structure of the standard loyalty field group might not be able to capture all relevant data for your particular loyalty program. In this scenario, you can choose to define a new custom field group to capture these fields instead.
Open the Add Field group dialog again, but this time select Create New Field group near the top. You are then asked to provide a display name and description for your field group.
As with class names, the field group name should be short and simple, describing what the field group will contribute to the schema. These too are unique, so you will not be able to reuse the name and must therefore ensure it is specific enough.
For this tutorial, name the new field group “Loyalty Details”.
Select Add field group to return to the Schema Editor. “Loyalty Details” should now appear under Field groups on the left-side of the canvas, but there are no fields associated with it yet and therefore no new fields appear under Structure.
Now that you have created the “Loyalty Details” field group, it is time to define the fields that the field group will contribute to the schema.
To begin, select the field group name in the Field groups section. Once you do this, the field group’s properties appear on the right-hand side of the editor and a plus (+) icon appears next to the name of the schema under Structure.
Select the plus (+) icon next to “Loyalty Members” to create a new node in the structure. This node (called
_tenantId in this example) represents your IMS Organization’s tenant ID, preceded by an underscore. The presence of the tenant ID indicates that the fields you are adding are contained in your organization’s namespace.
In other words, the fields you are adding are unique to your organization and are going to be saved in the Schema Registry in a specific area accessible only to your organization. Fields you define must always be added to your tenant namespace to prevent collisions with names from other standard classes, field groups, data types, and fields.
Inside that namespaced node is a “New Field”. This is the beginning of the “Loyalty Details” field group.
Using the controls on the right-hand side of the editor, start by creating a
loyalty field with type “Object” that will be used to hold your loyalty-related fields. When finished, select Apply.
The changes are applied and the newly created
loyalty object appears. Select the plus (+) icon next to the object to add additional loyalty-related fields. A “New Field” appears and the Field properties section is visible on the right-hand side of the canvas.
Each field requires the following information:
The first field for the
Loyalty object will be a string called
loyaltyId. When setting the new field’s type to “String”, the Field properties section becomes populated with several options for applying constraints, including default value, format, and maximum length.
Different constraint options are available depending on the data type selected. Since
loyaltyId will be an email address, select “email” from the Format dropdown menu. Select Apply to apply your changes.
Now that you have added the
loyaltyId field, you can add additional fields to capture loyalty-related information such as:
To add each field to the schema, select the plus (+) icon next to the
loyalty object and fill in the required information.
When complete, the Loyalty object will contain fields for loyalty ID, points, and member-since.
When defining fields in the Schema Editor, there are some additional options that you can apply to basic field types in order to provide further constraints on the data the field can contain. The use cases for these constrains are explained in the following table:
|Required||Indicates that the field is required for data ingestion. Any data uploaded to a dataset based on this schema that does not contain this field will fail upon ingestion.|
|Array||Indicates that the field contains an array of values, each with the data type specified. For example, using this constraint on a field with a data type of “String” specifies that the field will contain an array of strings.|
|Enum||Indicates that this field must contain one of the values from an enumerated list of possible values.|
|Identity||Indicates that this field is an identity field. More information regarding identity fields is provided later in this tutorial.|
|Relationship||While schema relationships can be inferred through the use of the union schema and Real-time Customer Profile, this only applies to schemas that share the same class. The Relationship constraint indicates that this field references the primary identity of a schema based on a different class, implying a relationship between the two schemas. See the tutorial on defining a relationship for more information.|
Any required, identity, or relationship fields are shown in the left rail, allowing you to locate these fields easily regardless of the schema’s complexity.
For this tutorial, the “loyalty” object in the schema requires a new enum field that describes the “loyalty level” of a customer, where the value can only be one of four possible options. To add this field to the schema, select the plus (+) icon beside the
loyalty object and fill in the required fields for Field name and Display name. For Type, select “String”.
Additional checkboxes appear for the field after its type has been selected, including checkboxes for Array, Enum, and Identity.
Select the Enum checkbox to open the Enum values section below. Here you can input the Value (in camelCase) and Label (an optional, reader-friendly name in Title Case) for each acceptable loyalty level.
When you have completed all field properties, select Apply to add the “loyaltyLevel” field to the
loyalty object now contains several loyalty-specific fields, and represents a common data structure that could be useful in other schemas. The Schema Editor allows you to readily apply reusable multi-field objects by converting the structure of those objects into data types.
Data types allow for the consistent use of multi-field structures and provide more flexibility than a field group because they can be used anywhere within a schema. This is done by setting the field’s Type value to that of any data type defined in the Schema Registry.
To convert the
loyalty object to a data type, select the
loyalty field under Structure, then select Convert to new data type on the right-hand side of the editor under Field properties. A green popover appears, confirming that the object has been successfully converted.
Now, when you look under Structure, you can see that the
loyalty field has a data type of “Loyalty” and the fields have small lock icons beside them, indicating they are no longer individual fields but rather part of a multi-field data type.
In a future schema, you could now assign a field as a “Loyalty” type and it would automatically include fields for ID, loyalty level, member since, and points.
You can also create and edit custom data types independently from editing schemas. See the guide on creating and editing data types for more information.
Your schema now contains several field groups in addition to the fields provided by its base class. When working with larger schemas, you can select the checkboxes next to field group names in the left rail to filter the displayed fields to only those provided by the field groups you are interested in.
If you are looking for a specific field in your schema, you can also use the search bar to filter displayed fields by name, regardless of which field group they are provided under.
The search function takes any selected field group filters into account when displaying matching fields. If a search query is not displaying the results you expect, you may need to double-check that you are not filtering out any relevant field groups.
The standard data structure that schemas provide can be leveraged to identify data belonging to the same individual across multiple sources, allowing for various downstream use cases such as segmentation, reporting, data science analysis, and more. In order to stitch data based on individual identities, key fields must be marked as Identity fields within applicable schemas.
Experience Platform makes it easy to denote an identity field through the use of an Identity checkbox in the Schema Editor. However, you must determine which field is the best candidate to use as an identity, based on the nature of your data.
For example, there may be thousands of loyalty program members belonging to the same “loyalty level”, but each member of the loyalty program has a unique
loyaltyId (which in this instance is the individual member’s email address). The fact that
loyaltyId is a unique identifier for each member makes it a good candidate for an identity field, whereas
loyaltyLevel is not.
The steps outlined below cover how to add an identity descriptor to an existing schema field. As an alternative to defining identity fields within the structure of the schema itself, you can also use an
identityMap field to contain identity information instead.
If you plan on using
identityMap, keep in mind that it will override any primary identity you add to the schema directly. See the section on
identityMap in the basics of schema composition guide for more information.
In the Structure section of the editor, select the
loyaltyId field and the Identity checkbox appears under Field properties. Check the box and the option to set this as the Primary identity appears. Select this box as well.
Each schema may contain only one primary identity field. Once a schema field has been set as the primary identity, you will receive an error message if you later attempt to set another identity field in the schema as the primary.
Next, you must provide an Identity namespace from the list of pre-defined namespaces in the dropdown. Since
loyaltyId is the customer’s email address, select “Email” from the dropdown. Select Apply to confirm the updates to the
For a list of standard namespaces and their definitions, see the Identity Service documentation.
After applying the change, the icon for
loyaltyId shows a fingerprint symbol, indicating that it is now an identity field.
Now all data ingested into the
loyaltyId field will be used to help identify that individual and stitch together a single view of that customer. To learn more about working with identities in Experience Platform, please review the Identity Service documentation.
Real-time Customer Profile leverages identity data in Experience Platform to provide a holistic view of each individual customer. The service builds robust, 360° profiles of customer attributes as well as timestamped accounts of every interaction customers have had across any system integrated with Experience Platform.
In order for a schema to be enabled for use with Real-time Customer Profile, it must have a primary identity defined. You will receive an error message if you attempt to enable a schema without first defining a primary identity.
To enable the “Loyalty Members” schema for use in Profile, begin by selecting “Loyalty Members” in the Structure section of the editor.
On the right-hand side of the editor, information is shown about the schema including its display name, description, and type. In addition to this information, there is a Profile toggle button.
Select Profile and a popover appears, asking you to confirm that you wish to enable the schema for Profile.
Once a schema has been enabled for Real-time Customer Profile and saved, it cannot be disabled.
Select Enable to confirm your choice. You can select the Profile toggle again to disable the schema if you wish, but once the schema has been saved while Profile is enabled, it can no longer be disabled.
Now that you have finished composing the schema, you can see the complete schema in the canvas. Select Save and the schema will be saved to the Schema Library, making it accessible by the Schema Registry.
Your new schema can now be used to ingest data into Platform. Remember that once the schema has been used to ingest data, only additive changes may be made. See the basics of schema composition for more information on schema versioning.
You can now follow the tutorial on defining a schema relationship in the UI to add a new relationship field to the “Loyalty Members” schema.
The “Loyalty Members” schema is also available to be viewed and managed using the Schema Registry API. To begin working with the API, start by reading the Schema Registry API developer guide.
The Platform UI shown in the following videos are out of date. Please refer to the documentation above for the latest UI screenshots and functionality.
The following video shows how to create a simple schema in the Platform UI.
The following video is intended to reinforce your understanding of working with field groups and classes.
The following sections provide addition information information regarding the use of the Schema Editor.
Experience Platform provides the flexibility to define a schema based on a class that is unique to your organization. To learn how to create a new class, see the guide on creating and editing classes in the UI.
You can change the class of a schema at any point during the initial composition process before the schema has been saved.
Reassigning the class for a schema should be done with extreme caution. Field groups are only compatible with certain classes, and therefore changing the class will reset the canvas and any fields you have added.
To learn how to change the class of a schema, see the guide on managing schemas in the UI.