MAP dataType in AEP Schema

[Pradeep-Jaiswal]

Has anyone tried MAP dataType for cased other than IdentityMap ? There is no example of data on adobe documentation. Usecase is ingested the ArrayOfObject data into MAP then create audience based on  specific index on the ArrayOfObject.

This is the sample data, which I want to try ingesting into MAP dataType

[

    {

        "date": "2024-12-29",

        "modelName": "PRP_TRIGGER_1Y_14DLAG",

        "model_score": "0.13705813884735107",

    },

    {

        "date": "2025-03-15",

        "modelName": "PRP_60D_LAPS_1Y_REV1",

        "model_score": "0.65977203845977783",

    }

]

[Tof_Jossic · Accepted solution]

@Pradeep-Jaiswal Found this documentation referring to Map fields in case that helps:

https://experienceleague.adobe.com/en/docs/experience-platform/xdm/ui/fields/map 

[GigiCotruta]

Hi @Pradeep-Jaiswal,

The problems that I see with the Map object that @Tof_Jossic sent the documentation are the restrictions:

• Map types MUST be of type object. --> No problem
• Map types MUST NOT have properties defined (in other words, they define “empty” objects). --> No problem
• Map types MUST include an additionalProperties.type field that describes the values that may be placed within the map, either string or integer. --> From your example, that's a problem as you defined a timestamp date value.
• Multi-entity segmentation can only be defined based on the map keys and not the values. --> That's a big problem IMO. I suppose you want to do your segmentation based on values, not keys. That's the main reason on why I would suggest you a normal object, not a map one. It does not have the limitations that this does.
• Maps are not supported for account audiences. --> I don't know if it is your case.

[Sukrity_Wadhwa]

Hi @Pradeep-Jaiswal,

Were you able to resolve this query with the help of the provided solutions, or do you still need further assistance? Please let us know. If any of the answers were helpful in moving you closer to a resolution, even partially, we encourage you to mark the one that helped the most as the 'Correct Reply.'

Thank you!