Course Practicalities

Glossary

This glossary is meant as a practical reading aid. The definitions are intentionally short and course-oriented, so they are easier to use while studying the chapters.

A

ACID — A common summary of transaction goals: atomicity, consistency, isolation, and durability.
Aggregate Function — A SQL function such as COUNT, SUM, or AVG that summarizes several rows into a smaller result.
Alias — A temporary name given to a column or table inside a query.
Anomaly — An unwanted situation caused by poor schema structure, such as having to update the same fact in many places or losing useful data when deleting one row.
Attribute — A named property in a data model. In an ER design it describes an entity, and in the relational model it corresponds to a column.
Auditing — Recording or inspecting who changed data and when.
Authorization — Rules controlling who is allowed to read or modify data.

C

Candidate Key — A column or set of columns that could uniquely identify a row, even if it is not the primary key chosen in the schema.
Cardinality — A description of how many instances of one entity can be related to another.
Column — A named field in a table that stores one kind of value for each row.
Constraint — A rule enforced by the database, such as NOT NULL, UNIQUE, CHECK, or a foreign key.
CRUD — Create, read, update, and delete: the four basic data operations.
CTE (Common Table Expression) — A named intermediate query introduced with WITH.

D

Data — Recorded facts such as text, numbers, timestamps, measurements, or other stored observations.
Data Lifecycle — The path data follows from creation through storage and use to archival, anonymization, or deletion.
Data Minimization — The idea that a system should store only the data that is actually needed for its purpose.
Data Independence — The idea that some changes at one abstraction level do not force changes everywhere else.
Data Type — A description of what kind of values a column can contain, such as integer, text, or timestamp.
Document Database — A database that stores structured documents, often in JSON-like form.
Database — The stored data itself.
DBMS (Database Management System) — The software that stores, retrieves, protects, and manages the data.
Derived Data — Data that can be computed from other stored data instead of being permanently stored as its own field.

E

Entity — A meaningful thing in the domain being modeled, such as a user, course, or exercise.
ER Diagram — A diagram showing entities, attributes, and relationships in a conceptual design.

F

Foreign Key — A column whose values refer to rows in another table.
Functional Dependency — A relationship where some attributes determine some other attributes.

G

Graph Database — A database designed around nodes and their connections.
Group / Grouped Result — A collection of rows treated together by GROUP BY.
GROUP BY — A SQL clause that partitions rows into groups so that aggregate functions can be computed per group.

H

HAVING — A SQL clause used to filter grouped results after grouping has taken place.

I

Index — A supporting structure that helps the DBMS reach some rows faster.
Information — Data that has been organized so that it supports interpretation.
Inner Join — A join that keeps only rows that find matches on both sides.
Isolation — The transaction goal that concurrent work should not interfere in unacceptable ways.

J

Join — A query operation that combines related rows from several tables.
Join Path — The sequence of table relationships that a query follows when it combines data from several tables.

K

Key-Value Database — A database that stores values behind keys and is especially useful for direct lookup by identifier.
Knowledge — A broader conclusion or understanding that someone draws from interpreted information.

L

Left Join — An outer join that keeps all rows from the left side even when no matching row exists on the right side.
Linking Table — A table whose main purpose is to store relationships between rows in two other tables, especially in a many-to-many design.

M

Many-to-Many Relationship — A relationship where one row on each side may be connected to several rows on the other side. In relational schemas this is usually represented through a linking table.
Migration — A tracked, reproducible change to the database schema.

N

Normalization — Improving schema quality by reducing redundancy and anomalies.
NULL — A marker meaning that a value is missing, unknown, or not applicable.

O

One-to-Many Relationship — A relationship where one row on one side can be connected to many rows on the other side, while each row on the many side points back to one row on the one side.
Operational Use — Using stored data to support the day-to-day actions of a running system.
Outer Join — A join that can keep rows even when no match is found on one side.

P

Parameterized Query — A query where values are passed separately from the SQL text instead of concatenated into it.
Personal Data — Data that relates to an identifiable person, either directly or when combined with other data.
PII (Personally Identifiable Information) — A practical term for data that can directly identify a person, such as a name, email, or phone number.
Primary Key — A column or set of columns that uniquely identifies a row in a table.
Privacy by Design — The idea that privacy should be considered already when deciding what to store, who may access it, and how long data should be kept.
Projection — In relational algebra, choosing columns.
Pseudonymized Data — Data where direct identifiers have been removed or replaced, but a person may still be identifiable indirectly.

Q

Query Plan — The DBMS’s chosen strategy for executing a query.

R

Relational Model — A way of structuring data using relations (tables), keys, and well-defined operations.
Relation — A table-like structure in the relational model.
Relationship — A connection between entities, such as one course having many exercises or one card having many tags.
Referential Integrity — The idea that references between tables should stay valid. For example, a foreign key should not point to a row that does not exist.
Retention — A decision or policy about how long data should be kept before it is deleted, archived, or anonymized.
Reporting Use — Using stored data to summarize, analyze, or review activity rather than to run day-to-day operations.
Row — A stored record in a table.

S

Schema — The structure of the database: tables, columns, keys, and constraints.
Seed Data — Starter rows inserted so that the application or tests have meaningful data immediately.
Selection — In relational algebra, choosing rows.
Serializability — The ideal that concurrent transactions should behave like some sensible serial order.
Set Operation — A query operation such as UNION, INTERSECT, or EXCEPT that combines or compares whole result sets.
SQL (Structured Query Language) — The language used to define, retrieve, and modify data in relational databases.
Subquery — A query written inside another query.

T

Table — The usual practical presentation of a relation as rows and columns.
Time-Series Database — A database designed for values or events that are primarily organized by time.
Transaction — A unit of work that should succeed or fail as a whole.
Transaction Boundary — The point that determines which database actions belong together in one transaction.
Tuple — The formal relational-model term for a row.

U

UNION — A SQL set operation that combines compatible result sets into one result. Duplicate rows are removed unless UNION ALL is used.

V

Vector Database — A database designed for storing vectors and finding items that are similar in meaning or content.

W

Walking Skeleton — A minimal end-to-end version of the project that already runs through all the main layers.
Wide-Column Database — A database that groups data by a row key and can store many columns, often with sparse structure.
WHERE — A SQL clause used to filter rows before grouping or final output.

← Project Roadmap

SQL Quick Reference →