Lecture ReVISION part 4

Lecture ReVISION part 4

Page 1: Introduction

• Title: BSc (Hons) ORYX UNIVERSAL COLLEGE

• Program: Computer Science

• Partnership: Liverpool John Moores University

• Presenter: Dr. Ali Baydoun

Page 2: Relational Databases

• Data is organized into tables (relations) consisting of rows and columns.

• Rows: Identified by a unique key (primary key).

• Columns: Represent attributes that provide values for the rows.

• Each table typically represents one entity type (e.g., customer, product).

• Rows are instances of the entity type, while columns store specific details (e.g., name, price).

Page 3: Getting Data Out of a Relational Database (SQL)

• Data extraction is performed using Structured Query Language (SQL).

• SQL is designed to manage structured data that includes relationships among entities and variables.

Page 4: Relational Database Example

• Relational Constraints: Primary Keys (PKs) and Foreign Keys (FKs).

• Important principles include ACID properties (Atomicity, Consistency, Isolation, Durability).

• Transactions can be cascaded.

• Queries are often relational and JOIN-based.

Page 5: Issues with Relational Databases

• Scalability: Poor performance across multiple servers.

• Challenges with resource elasticity.

• Difficulty in modeling certain data types without major restructuring (Normalization required).

• Databases may be ineffective or inefficient.

Page 6: Shortcomings of the Relational Model

• Relational approaches are becoming limited due to the massive increase in data production.

• Recent advances aim to address these limitations, but challenges remain with large datasets needing distributed databases.

• The emergence of diverse data types (tweets, video, etc.) makes structured schemas inadequate.

• This has led to the development of NoSQL database approaches.

Page 7: NoSQL Databases

• NoSQL stands for "NOT SQL" or "Not Only SQL."

• Characterized by a flexible approach to data storage suitable for handling "big data."

• Various types of NoSQL databases:

  • Key-Value

  • Graph

  • Column

  • Document

Page 8: NoSQL Database (Key-Value)

• Flexible Structure: Key-value databases are an alternative to relational databases, lacking predefined schemas.

• Values: Can be a wide range of data types (numbers, strings, JSON, etc.).

• Functionality: Accessed through unique keys; no query language exists.

• Operations: Basic functions include:

  • Put: Adds/updates key-value pairs.

  • Get: Retrieves values for given keys.

  • Delete: Removes key-value pairs.

Page 9: Key-Value Database Functionality

• Data Types: Can store any data type including BLOBs.

• Similar to a dictionary with unique keys indexing values.

• Enables rapid retrieval of values regardless of database size.

Page 10: Key-Value vs. RDBMS

• Key-value databases treat values as single opaque collections with varying fields.

• Provides flexibility similar to object-oriented programming.

• Memory-efficient storage leading to performance gains.

Page 11: NoSQL Databases - Graph

• Graph Databases: Use graph theory for storing and querying relationships.

• Comprised of nodes (entities) and edges (connections).

• Unique identifiers and properties define each node.

Page 12: Applications of Graph Databases

• Suited for analyzing connections, especially in social media data mining.

• Useful in business data that involves complex relationships (e.g., supply chain management).

• Concept attributed to mathematician Leonhard Euler.

Page 13: NoSQL Databases - Document

• Document Databases: Store data in JSON-like documents.

• Facilitates data storage that mirrors application code, flexible and semi-structured.

• Adaptable to application evolution, supporting catalogs, user profiles, etc.

Page 14: Document Database Uses

• Ideal for Content Management Systems (e.g., blogs, video platforms).

• Allows for easy updates and changes without schema updates or downtime.

Page 15: Document Database for Catalogs

• Efficient for storing product catalogs; handles diverse attributes without relational inefficiency.

• Each product can be described in a single document, enabling easier management and reading.

Page 16: Unstructured and Semi-Structured Data

• Flexible Design: Reduces schema design requirements.

• Encourages reduced constraints on database querying, simplifying access.

Page 17: Who Uses NoSQL Technologies?

• Massive web applications (e.g., Amazon) catering to large user bases.

• Big Data Applications: Platforms like Twitter and Facebook manage high-frequency updates and data distribution.

• Content Management Systems: NoSQL databases provide flexibility for managing diverse content types and structures, enabling rapid development and deployment.

create flash cards