Consider the following JSON snippet:

{  
   "id":101,
   "name":"John Doe",
   "nickname":"John",
  "username":"johndoe",
   "email" : "[email protected]",
   "createdAt":"2018-08020T10:10:00",
   "friendIds":[  
      2,
      3,
      4,
      10
   ],
   "employement":[  
      {  
         "organizationName":"SAS Web Tech",
         "startDate":"2017-08-06"
      },
      {  
         "organizationName":"Kirlosker Service Center",
         "startDate":"2016-08-06"
      }
   ],
   "gender":"Male"
}

The JSON data holds information of an employee who has a name, nickname, and username, and the date when the record was inserted. An employee can be associated with one or more organizations. To create a schema in AsterixDB, our schema could appear as follows:

create type EmploymentType as closed {
  organizationName: string,
  startDate: date,
  endDate: date?
};

Notice that the schema is defined as closed. Now we can define UserType as follows:

create type UserType as open {
  id: int,
  username: string,
  name: string,
  email: string,
  nickname: string,
  createdAt: datetime,
  friendIds: {{ int }},
  employment: [EmploymentType],
  gender: string
};

The world of data in AsterixDB is organized into data namespaces called dataverses. To set default dataverses for a series of statements, the use dataverse statement is provided. For example, to use dataverse to be TinySocial, we need to execute the following statement:

$ use dataverse TinySocial;

Now create dataset is used to create a new dataset. To create a dataset for this example, we need to run the following statements:

$ create dataset Employee(UserType) primary key id;

After that, we can create the example UserType and EmploymentType. When you run these statements, the output screen should look something like this: