Types of NoSQL Databases

This article provides an overview of the different types of NoSQL databases - Key Value Store, Document Store, Column Store, and Graph Base. It explains how each type works and provides examples of real-world applications.

Introduction:
NoSQL database which also referred as Not Only SQL. This is considered as new approach towards
design and management of database having very large set of data. NoSQL overcomes the disadvantages
of Relational Database such as scalability and performance issues. The data in NoSQL database is
distributed over multiple virtual cloud servers and can be scaled whenever there is a requirement for
scale. NoSQL is more useful for the enterprises which deals with large amount of unstructured data as
NoSQL does not have restriction over schema. There are in total four categories in NoSQL which are
Key-Value Store, Document-based Store, Column-based Store, and Graph-based. These all four
categories are explained in details in below sections (Sullivan, 2015).
Key Value Store NoSQL Database:
Key value store is considered as the simplest type of NoSQL for storing data. In this type of NoSQL the
data is stored in form of Key-value pair. Here it uses hash table which has unique keys and pointers
corresponding to each data which is stored. The client can perform operation such as fetch data
corresponding to key or assign a key for the value or delete the key from the database. The data stored
in the form blob and it is the responsibility of the application to understand what kind of data is stored.
The implementation of it is very easy and has high scalability and great performance due to its Key-Value
nature. Examples of this database are MemcahceDB, Redis and Riak (Kumar, n.d).
Application of these database in real word-
 These are used for storing session data for user.
 Maintain schema-less profiles of user.
 Store preferences of user
 Storing shopping cart data.
Example:

Document store NoSQL database
Document store NoSQL is somewhat similar to key value databases as in this type as well there is a
relationship of key and value. Similarity is that for both the data is identified with the help of key and the
difference is that in document database the value is either structured or semi structured data and these
document (data is referred as document) can be in JSON, BSON or XML format. Here in this type of
database Key uniquely identifies the document (which has data). The examples of this kind of database
is ElasticSearch, MongoDB and Apache CouchDB (Kumar, n.d).
Application of these database in real word-
 Widely used on E-commerce platforms
 Content management systems
 Analytics platforms
 Blogging platforms
Example:
Column store NoSQL database:
Column Store NoSQL database is different from the above specified databases as it does not have key
value pair instead it stores the data in cells which are grouped in columns. The nature of this data base is
column-oriented and columns are grouped into families of column. The family of column has unlimited
number of column which can be created while defining the schema or at runtime. The operation are not
performed at row but at columns. The advantage of using this database is due to its fast search/access
and data aggregation. Examples of this type of databases is- Cassandra and Apache Hadoop Hbase
(Sullivan, 2015).
Application of these database in real word-
 Content management systems
 Blogging platforms
 Systems that maintain counters
 Services that have expiring usage
 Systems that require heavy write requests.

Example:
The 2 dimensional below RDBMS row representation in Column database:
In Column Database:
Graph base NoSQL database
Last category of NoSQL database is Graph base. This type is completely based on Entity- Attribute-Value
model. The entities are also called as Nodes and nodes stores the data and relationships describes the
relationship between the nodes. The advantage of Graph database is that it allows to add any
relationship virtually on the fly. Examples of this databases are Neo4j, ArangoDB and OrientDB.
Application of these database in real word-
 This database is widely used in Fraud detection
 Graph based search
 Network and IT operations
 Social networks
Example:

References:
Kumar, G. (n.d).Exploring the Different Types of NoSQL Databases. Retrieved from:
https://www.3pillarglobal.com/insights/exploring-the-different-types-of-nosql-databases
Sullivan, D. (2015).Types of NoSQL databases and key criteria for choosing them. Retrieved from:
https://searchdatamanagement.techtarget.com/feature/Key-criteria-for-choosing-different-types-of-
NoSQL-databases