Data Models

In the proposed database model, the emphasis is on leveraging NoSQL databases like Firebase Firestore due to their inherent suitability for key-value storage and scalability. Firebase Firestore employs partition keys to shard or distribute records across the database, promoting efficient data distribution and retrieval.

The uniqueness of message IDs is confined within the scope of partition keys, which means that a message ID is not globally unique but rather unique within its partition. This design choice aligns with the distributed nature of NoSQL databases and ensures efficient data organization

Example Firestore Data Models

Users Collection

{
  "users": {
    "userId1": {
      "username": "johndoe",
      "email": "johndoe@example.com",
      "firebaseRole": "pro",
      "subscription_id": "sub_123456789",
      "language_from_id": "en",
      "language_to_id": "fr"
    },
    "userId2": {
      "username": "janedoe",
      "email": "janedoe@example.com",
      "firebaseRole": "non-pro",
      "subscription_id": "sub_987654321",
      "language_from_id": "en",
      "language_to_id": "es"
    }
  }
}

The role of firebaseRole within the subscription_id field introduces a mechanism for categorizing users into pro and non-pro members. This information is vital for determining subscription status and privileges within the system.

The involvement of language_to_id and language_from_id points to a language translation feature facilitated by the Google Cloud Artificial Intelligence Service. These fields likely store identifiers linking to specific languages, enabling seamless translation services within the application.

To manage relationships between users and groups, two tables are suggested. The Group Membership table is designed to support message broadcasting, a crucial functionality that involves determining the intended recipients of a message within specific user groups.