Data Modelling

    Avin Khair

    To create a database, a special database is needed, it needs the desired database structure, the desired relationships, the desired target and usefulness as well. There are many types of modeling and types from specific databases.

    1. Relational
    Relational data requires more than 1 table, where each table must have a main id used to connect between tables (join)
    2. Dimensional
    Actually it is almost the same as relational, only the dimensional data is focused on one table in which there are various primary keys from each table, which later will be connected to each other.
    3. Graph
    This data model has a freer relation where, not affected by the table structure must be followed. This model is based on nodes and edges. Where the node itself is the entity and edge is the relation or path that connects between entities freely. The relation of this model has no restrictions.
    4. Document
    This new database model consists of objects. Where we can store multiple data. We can analogize documents as a table where we can enter another table in the table value.
    5. Wide Column
    This model consists of only one table, fixed table and does not have a join. So we can say that all data is combined, but in a single table.
    6. Key Value
    Models that have predetermined data, so that the new data to be entered must be in accordance with the set value rules.

    Queries are language or systematic that are used to manipulate data. This data manipulation can be filtering, join data, delete, update, insert or display data. Queries can be used on 2 database systems namely sql and nosql.

    SQL or Structured Query Language is a language used to access data in a relational database. This language is de facto the standard language used in relational database management. So sql can only be used for relational, and dimensional data models.

    NoSQL is not a language. NoSQL is a mechanism for storing data and retrieving data carried out by our database. NoSQL does not require relational data models and SQL language to do this. NoSQL uses metadata in our database and uses the index of that data. NoSQL has four mechanisms:
    · Table-oriented, for example: Google with Big Table and Facebook with Cassandra,
    · Graph-oriented,
    · Document-oriented databases, for example: MongoDB and also CouchDB,
    · Key-value store, for example: Memcache and Redis

    Leave a comment

    Your email address will not be published.