mongodb multiple collection join with groupby

I am noobie in mongodb,
<br>but I am trying to aggregate data based on type of document(type_id)
I have following collections

items<br>

{
    "_id" : "BE05332051E0AD282",
    "type_id" : "684EB00155D031225",//Organization or Book or Person
    "name" : "Owi Student Union",
    "structure" : {
        "organization_active_status" : "Inactive",
        "country" : "USA"
    },
    "hash" : "59E48A8C3F7B20",
    "geometry" : {
        "type" : "Point",
        "coordinates" : [
            -85.48172999999997,
            32.60746000000006
        ]
    },
    "geometry_confidence" : 0.0,
    "has_geometry" : true,
    "is_node" : false,
    "groupids" : [
        "DE05332041E0A4978"
]
}

tuples<br>

{
    "_id" : "0E05333051E0A02E7",
    "source_id" : "BE05332051E0AD282",
    "destination_id" : "9E05332051E0AF344",
    "edge_id" : "DE05332051E0A3B5B",
    "direction" : NumberInt(1),
    "changed" : true
}
{
    "_id" : "0E05333061E0A02E8",
    "source_id" : "BE05332051E0AD282",
    "destination_id" : "0E05333051E0A02E7",
    "edge_id" : "DE05332051E0A3B5B",
    "direction" : NumberInt(1),
    "changed" : true
}

types<br>

{
    "_id" : "684EB00155D031225",
    "name" : "Organization"
},
{
    "_id" : "BE0532A041E0AC077",
    "name" : "Book"
}

1. items collection contain all type of document for organization, books and can be differentiate on type_id
2. each organizations may have multiple books, person

relations

1) tuples.source_id = items._id
2) tuples.destination_id = items._id
3) items.type_id = type._id

I am trying aggregate data like this, but unable to get complete result

db.tuples.aggregate([
    {
        $lookup:{
            from: "items",
            localField: "destination_id",
            foreignField: "_id",
            as: "source" 
        }
    },
    {
        $match:{
           "source_id":"5BD5A377499D406EE05332051E0A93C2",
           "source.type_id" : "43B369EDE96A4876E0532A041E0A995A"
        }
    }
]);

Could any one help me on how can I create final result like this

Organizations

{
    "_id" : "BE05332051E0AD282",
    "type_id" : "684EB00155D031225",//Organization
    chapters : {
     "list of chapter(documents from item collection)"
    },
    books : { 
    "list of books(documents from item collection)"
    }
}

I would really appreciate any kind of help or suggestion, Thanks

You can achieve this with aggregation pipeline as below. You can run it here,
https://mongoplayground.net/p/H2Ygqgjp2f5

    db.items.aggregate([
  {
    "$match": {
      "type_id": "684EB00155D031225"
    }
  },
  {
    "$lookup": {
      from: "tuples",
      localField: "_id",
      foreignField: "source_id",
      as: "tuples"
    }
  },
  {
    "$lookup": {
      from: "items",
      let: {
        destination_ids: "$tuples.destination_id"
      },
      pipeline: [
        {
          $match: {
            $expr: {
              $in: [
                "$_id",
                "$$destination_ids"
              ]
            },
            "type_id": "BE0532A041E0AC077"
          }
        }
      ],
      as: "books"
    }
  },
  {
    "$lookup": {
      from: "items",
      let: {
        destination_ids: "$tuples.destination_id"
      },
      pipeline: [
        {
          $match: {
            $expr: {
              $in: [
                "$_id",
                "$$destination_ids"
              ]
            },
            "type_id": "BE0532A041E0AC087"
          }
        }
      ],
      as: "persons"
    }
  },
  {
    "$project": {
      "_id": 1,
      "type_id": 1,
      "name": 1,
      "books": 1,
      "persons": 1
    }
  }
])

this would return a result as this.

  [
  {
    "_id": "BE05332051E0AD282",
    "books": [
      {
        "_id": "9E05332051E0AF344",
        "name": "Test Book 1",
        "type_id": "BE0532A041E0AC077"
      }
    ],
    "name": "Owi Student Union",
    "persons": [
      {
        "_id": "0E05333051E0A02E7",
        "name": "Test Person 1",
        "type_id": "BE0532A041E0AC087"
      }
    ],
    "type_id": "684EB00155D031225"
  }
]

But I would suggest you rethink your database design, since you are using a nosql database you should be able to embed documents and have a tree like structure. Hope this helps.