Query inner object without knowing parent key path

Elastic Stack Elasticsearch
1

I have a items with below structure

{
    "user_id" : "unique id for the user",
    "minutes": {
        "2019-1-11T14:41:00.000Z": {
            "interest_1" : {
                    "count" : 2,
                    "obs" : [
                            "req_id_1","req_id_2"
                    ]
            },
            "date" : "2019-1-11T14:41:00.000Z"
        },
        "2019-1-15T12:18:00.000Z": {
            "interest_1" : {
                    "count" : 1,
                    "obs" : [
                            "req_id_3"
                    ]
            },
            "interest_2" : {
                    "count" : 1,
                    "obs" : [
                            "req_id_4"
                    ]
            },
            "date" : "2019-1-15T12:18:00.000Z"
        },
        "2019-1-15T13:27:00.000Z": {
            "interest_1" : {
                    "count" : 3,
                    "obs" : [
                            "req_id_5","req_id_6","req_id_7"
                    ]
            },
            "date" : "2019-1-15T13:27:00.000Z"
        },
        "2019-1-15T20:30:00.000Z": {
            "interest_2" : {
                    "count" : 1,
                    "obs" : [
                            "req_id_8"
                    ]
            },
            "date" : "2019-1-15T20:30:00.000Z"
        }
    }
}

I want to get all users with interest count and observations from a list of interests between a time interval.

For example, i want all users with interest interest_1 between time 2019-1-15T12:00:00.000Z and 2019-1-15T14:00:00.000Z

I am using this structure to track time series event of the user. I know elasticsearch allow inner objects query but we have to know the whole path to inner object from root object. But in my case, i cannot say the parent object key.

What will be the query to get my result or is there any other good method to track time series?

2

That is a very bad document structure to use as the number of fields in your mappings will just grow and grow over time as you get a new one every minute. I would recommend changing how you structure your data, which will also change how you query it.

3

Ok. I will figure out the structure.
I need to know how to query on nested key without knowing the parent key path. Is it possible in ElasticSearch?

4

I have just recommended that you do not use dynamically generated field names like in your example, which means that you would know the key paths.