Database Types and Structures

Vocabulary terms covering different types of database systems including relational, graph, column stores, and NoSQL variants like key-value and document databases.

Relational Databases

Databases where primary keys prevent duplications, foreign keys enforce referential integrity, and normal rules prevent anomalies to ensure data consistency.

Primary keys

Keys in a relational database used to prevent duplications.

Foreign key

A key that enforces referential integrity within a relational database.

Normal rules

Guidelines in relational databases used to prevent anomalies.

Graph database

A type of database viewed as a mathematical graph where the relationship between the data is the key information.

• Can be part of larger system

• Doesn’t have joins so can scale easily

• Has issues when it gets larger with scaling

Column stores

A database type similar to relational databases but data is stored in columns not rows, making searching by column much faster.

• Good when data doesn’t change

• Searching for rows is hard, complicates WHERE clause

• Adding rows is difficult

Hashing (Numerical storing)

A process where a simple key is mapped to another more complex value to determine its storage location.

• Quick access due to array like storing

Hash function

A function applied to a value to determine the specific storage location for it.

Collisions

Issues that arise in hashing due to a poor hash function where multiple keys map to the same storage location.

Access speed of hash tables

$O(1)$ access to elements.

Key-value stores

A NoSQL database that acts like a dictionary or an array of keys where values are the records in the table.

• Keys need to be unique otherwise prevent collisions

• If data is variable relational would have many NULL values

Disadvantages of hashing

• lack of referential integrity

• Basic querying only

• Only search for exact key

Document database

A self-contained database using a document-oriented model similar to key-value stores but allowing for more complex queries and returning specific parts of a document.

CREATE, READ, UPDATE, DELETE

The standard operations (CRUD) used in document databases.

JSON

A semi-structured data format formed of key, value pairs used in key-value and document databases.

BSON

A data storage format that is not human readable, has more data types than JSON, and can be processed faster than JSON.