IB Comp Sci Option A Databases HL

Relational

Stores data in tables with relationships between them.

Object-oriented

Stores data as objects, similar to object-oriented programming languages.

Network

Based on a traditional hierarchical database, Data is represented using a graph, with records and links. In Networsk Child Records can have multiple parent records

Spatial

A Type of Databse Optimized for storing and querying data representing objects defined in a geometric space.

Multi-dimensional

Data is organized into dimensions, often used in data warehousing and OLAP (Online Analytical Processing)

Data Definition

The process of specifying the structure of a database — including the tables, fields (columns), data types, and relationships.

Data Manipulation

The process of adding, changing, retrieving, or removing data in a database.

Data Integrity

The accuracy, consistency, and reliability of data in the database.

Data Warehouse

A subject-oriented, integrated, time-variant, and non-volatile collection of data is used in decision-making.

A large Centralized repository for storing integrated data from multiple sources.

Subject-oriented

Organized around major subjects like sales, product.

Integrated

Data is collected from various sources and merged into a coherent whole.

Time-variant

Data is kept for historical analysis and has a time dimension.

Non-volatile

Once entered into the warehouse, data is not updated.

Strategic Planning

Using data to make long-term decisions.

Business Modelling

Creating data models to simulate different business scenarios.

Time Dependency

Data in a warehouse is valid for a specific range of time.

Real-time Updates

Data is constantly refreshed from operational systems.

Advantages of Data Warhouse

Centralizes data management

Supports decision-making with complex queries.

Enables historical analysis and trend identification over time

Allows for faster query performance due to optimized structure (e.g., de-normalized schema)

Doesnt Effect Daily Operations

Extract

Pulling data from various sources.

Such as:

Operational Databases

CRM (Customer Relationship Management) System

ERP (Enterprise Resource Planning) system

Spreadsheets & CSV files

Transform

Extracted data is cleaned, restructured, and converted to match the format and rules of the target data warehouse.

Deduplicate

Clean (Remove Invalid or Missing Data)

Standardize (Make sure all data follows same rules)

Loading

Inserting data into the final target database or data warehouse.

Data Cleaning

Enhancing data quality by rectifying inconsistent, incomplete, or inaccurate data.

Pattern Discovery

Identifying patterns or correlations in large datasets using methods like cluster analysis, associations, classifications, sequential patterns, forecasting.

Benefits of Pattern Analysis

Fraud detection in banking

optimizing retail marketing strategies.

Disease Detection in Healthcare

Predictive Modelling

Using statistical techniques for prediction like decision tree induction and neural networks.

Database Segmentation

Dividing a database into distinct segments based on certain criteria.

Link Analysis

Analyzing connections between nodes in a network to identify relationships and patterns.

Deviation Detection

Identifying unexpected or rare items, events, or observations that raise suspicions by differing significantly.

Decision tree induction

A predictive modelling technique where a system learns from historical data by building a tree-like structure of decisions and outcomes.

Cluster Analysis

A method of grouping similar data points together based on shared characteristics, without pre-labeled categories.

• Uses Unsupervised learning

• Example: Segmenting customers into groups based on shopping habit

Associations

Rule Learning, an (If, Then)

Identifies relationships between variables by finding items that frequently occur together in a dataset.

• Often used in market basket analysis

• Example: “If a customer buys bread and butter, they are likely to buy jam”

Classifications

Assigns data into predefined categories or classes based on learned patterns from labeled training data.

• Uses Supervised learning

• Example: Predicting if an email is spam or not based on its contents

Sequential patterns

Identifies frequently occurring ordered sequences of events or items in a dataset.

• Focuses on time-based patterns

• Example: A user who buys a phone is likely to buy a case a week later, then a charger

forecasting

Predicts future values based on patterns found in historical data, using techniques like regression or time series analysis.

• Example: Stock Market Prices