Database Types

A set of vocabulary flashcards covering the characteristics, pros, and cons of Graph, Column, Key-Value, and Document databases as described in the lecture notes.

Relational Databases

Databases that use primary keys to prevent duplication, foreign keys to enforce referential integrity, and normal form rules to prevent anomalies, ensuring data is always consistent.

Graph Databases

A database type based on a mathematical graph where the key is made from the relationships between the data, often used in social media for relationship analysis.

Column Stores

A database model that uses tables but stores data in columns rather than rows, allowing rows to have varying numbers of columns.

Hashing

The process where a key is mapped to another value using a hash function to determine its storage location in a data structure called a hash table.

Key collisions

Problems that occur in key-value stores when a poor hash function results in multiple keys attempting to use the same storage location.

Key-value Stores

Databases that function as giant hash tables where unique keys are mapped to records with different fields, often providing $O(1)$ retrieval performance.

Key-Value Performance

Systems that can have a large number of records and still retrieve values quickly if the key is known, often achieving $O(1)$ consistency performance due to hashing.

Key-Value Operations

A limited set of basic commands consisting only of "PUT", "GET", and "DELETE" operations, which do not support complex queries.

Document Databases

A NoSQL database type that uses a document-oriented model to store self-contained data where each document may use different fields.

Document vs. Key-value Retrieval

In key-value stores "GET" returns the entire field, whereas in document databases it can return specific parts of the field.

Document Database Operations

Permits basic operations known as create, read, update, and delete, and can be accessed via programming languages.

Graph Databases PROS

Excellent for interconnected data. Doesn’t require a lot of joins, giving it a flexible schema.

Column Stores PROS

Efficient for analytical queries as its easy to search for colums, and to retrieve values

Hashing PROS

Fast data retrieval, efficient use of space.

Key-value Stores PROS

No need to ensure consistency due to hashing + allows a variety of data values that can be retrieved quickly with the key

Document Databases PROS

Flexible schema and able to store diverse data types.

Graph Database CONS

as graph gets bigger and more people join the network, it can slow down and become harder to scale

Column Stores CONS

Cons: Not ideal for transactional operations, as it’s hard to search for and add to rows

Hasing CONS

poor hash functions can cause key collisions.

Key-value stores CONS

lack of referential integrity means the different data values can lead to contradictions as mistakes aren’t being checked

Document database CONS

can only perform basic operations and has no normalisation