Summary Principles of Database
Management
Chapter 1: fundamental concepts of database management
Key definitions
Database = collection of related data items withing specific business process or
setting
Database Management System (DBMS) = software package used to define, create,
use and maintain data
Database system = combination of database and DBMS
File versus Database approach to data management
File approach
o Duplicate/redundant data
will be stored for each
application separately
o Danger of inconsistent data
o Hard to integrate applications
/ make changes
Database approach
o Better than file approach
o More efficient, consistent,
less storage needed
o Facilities provided for data
querying
1
,Elements of a database system
Database model versus instances
o Database model: provides description of database at different levels of detail
Specified during database design, don’t normally change
Stored in catalog
Example:
Student (number, name, email, …)
Course (number, name, location, …)
o Database state: represents data in the database at a particular moment
Also called current set of instances
Changes every time data gets updated / removed
Data model
o Database model: comprised of different data models
Each describing the data from different perspectives
o Conceptual data model: high-level description of data items
Used for communication between architect and user
Usually represented using Enhanced-Entity Relationship (EER) model
o Logical data model: translation of conceptual model towards a specific
implementation environment
Can be hierarchical, CODASYL, relational, OO, extended relation, XML
or NoSQL
Can be mapped to internal data model that represents data’s physical
storage details
Describes what’s stored where, what format, …
2
, Can only describe structured data
o External data model: contains subsets of data in logical model (=views),
specialized for specific applications for certain groups of users
The three layer architecture
Catalog
o Heart of the DBMS
o Contains data definitions/metadata of the database application
o Stores definitions of the views, logical and internal data models
o Synchronizes all models to guarantee consistency
Database users
o Information architect: designs conceptual data model
Thorough interaction with business user
o Database designer: translates conceptual model to logical/internal model
o Database administrator (DBA): implementation and monitoring of database
o Application developer: develops the actual applications in ex. Python
o Business user: runs these applications to perform specific operations
Database languages
o Data definition language (DDL): used by DBA to express internal, external and
logical data models
Definitions stored in the catalog
o Data manipulation language (DML): retrieve, insert, delete and modify data
3
, DML statements: can be embedded in programming language
o Structured query language (SQL): offers both DDL and DML statements for
relational database systems
Advantages of database systems and management
Data independence
o = changes in data definitions: little to no impact on applications
o Physical data independence: when changes are made to internal data model
(data storage):
Neither application, views, logical data model must be changed
DBMS: should provide interface between logical/internal data models
o Logical data independence = software applications are minimally affected by
changes in conceptual/logical data model
Views in external model: act as protective shield
DBMS: must provide interface between conceptual/logical/extended
layer
Database modelling
o Data model = explicit representation of data items, together with their
characteristics/relationships
o Conceptual data model: should provide mapping of the data requirements of
the business user
Translated into logical/internal data model
o Data model’s assumptions/shortcomings must be clearly documented
Managing structured, semi-structured and unstructured data
o Structured data (ex. Spreadsheet)
Can be described with logical data model
Ability to force integrity rules, correctness of data
Facilitates searching, processing and analysing the data
o Unstructured data (ex. Text, picture)
No components in a file that can be interpreted by a DBMS/application
Volume of unstructured data > volume of structured data
o Semi-structured data (ex. Emails, HTML)
Has certain structure, but irregular and highly volatile
4
