DCIT 24 Reviewer
DCIT 24 Reviewer
System (SMIS)
• Records/basic info
Database
• Records (academics)
• Storage • Medical Records
• Server • Enrollment
• Manage • Capability Enhancement records
• File compilation
Superadmin
• Report Generation
Admin
Goals: Good decision making requires good End users
information derived from raw facts
Data Vs Information Advantages of a DBMS:
• Collection of programs
• Manages structure and controls access
to data
Database location(s) • Database design focuses on design of
database structure used for end-user
1. Centralized Database
data
• Data located in a single location
– Designer must identify
2. Distributed Database database’s expected use
• Different sites/branches with each • Well-designed database:
other own database
– Facilitates data management
• Cloud database (main database)
– Generates accurate and valuable
3. Relational Database
information
• Rows/columns tables
• Poorly designed database:
• Compare data on database
– Causes difficult-to-trace errors
Two types of distributed database
Evolution of File System Data Processing
1. Operational Database (OLTP)
• Reasons for studying file systems:
• Online Transaction Processing
database – Complexity of database design
• extent of usage/data usage is easier to understand
• supports a company’s day-to-day
– Understanding file system
operations
problems helps to avoid
• Transactional or production database
problems with DBMS systems
2. Data Warehouse (OLAP)
– Knowledge of file system is
• Online Analytic Processing Database useful for converting file system
system to database system
• stores data used for tactical or strategic
• File systems typically composed of
decisions.
collection of file folders, each tagged
Homogenous Database and kept in cabinet
• Same on all databases – Organized by expected use
Heterogenous Database • Contents of each file folder are
logically related
• Different OS, hardware and software
• Manual file systems
MS Access
– Served as a data repository for
• Single and work
small data collections
• Centralized
– Cumbersome for large
Why Database Design Is Important
collections
• Computerized file systems
– Data processing (DP) specialist tasks for which spreadsheets are
converted computer file not appropriate – database
structure from manual system substitute.
• Wrote software that Problems with File System Data
managed the data Processing
• Designed the application • File systems were an improvement
programs over manual system
• Initially, computer file systems – File systems used for more than
resembled manual systems two decades
• As number of files increased, file – Understanding the shortcomings
systems evolved of file systems aids in
development of modern
– Each file used its own
databases
application program to store,
retrieve, and modify data – Many problems not unique to
file systems
– Each file was owned by
individual or department that • Even simple file system retrieval task
commissioned its creation required extensive programming
– Ad hoc queries impossible
– Changing existing structure
difficult
• Security features difficult to program
– Often omitted in file system
environments
• Summary of file system limitations:
File System Redux: Modern End-User
Productivity Tools – Requires extensive
programming
• Ubiquitous use of personal
productivity tools can introduce the – Cannot perform ad hoc queries
same problems as the old file systems – System administration is
• Microsoft Excel complex and difficult
3. External Level (Visual)- 60% even more The relational model is widely used in
abstract than the logical level. For end users relational database management systems
with lowest degree of detail. This is what an (RDBMS) and forms the basis for SQL
end-user gets to see. He/she does not get the databases.
entire database, but depending on the queries Referential Integrity Rule in Relational
made from the front-end the user gets to see Database Model
the data. It may be a single data from the
entire database or a collection of data in Referential Integrity Rule in DBMS is based
tabular format. Multiple views of the same on Primary and Foreign Key. The Rule
data are available to the user, the defines that a foreign key have a matching
representation can be a table, a graph, or a pie primary key. Reference from a table to
chart. View Level is the highest level of data another table should be valid.
abstraction in DBMS. Referential Integrity Rule examples
Lesson 3
Relational Database Model
& Entity Relationship Diagram
The rule states that the DEPT_ID in the
Employee table has a matching
Logical View of The Data valid DEPT_ID in the Department table.
A logical view of data refers to the To allow join, the referential integrity rule
representation of data without necessarily states that the Primary Key and Foreign Key
considering the physical implementation or have same data types.
storage details. It focuses on how data is
organized, structured, and related to each
other in a way that makes sense for the users
and the applications interacting with the data.
The logical view is often expressed through
data models, which define the relationships,
constraints, and rules governing the data.
Entity Integrity Rule • Note: Secondary Key is not a Foreign
Key.
• Make sure that each tuple in a table is
unique. Let’s say you selected Student_ID as
• Every table must have a primary key, primary key,
for example, Student_ID for a Student therefore Student_Enroll and Student_Em
table. ail will be Secondary Key (candidates of
• Every entity is unique. primary key).
• The relations Primary Key must have
unique values for each row.
• Primary Key cannot have NULL value
and must be unique.
• Example can be
an Employee_ID cannot be null in an C. SUPER KEY
Employee table. Super Key is an attribute (or a set of
Key attributes) uniquely identifying a tuple, i.e.,
an entity in an entity set.
A. COMPOSITE KEY
It is a superset of Candidate Key, since
• A primary key having two or more
Candidate Keys are selected from super key.
attributes is called composite key. It is
a combination of two or more columns. Let us see an example −
• An example can be −
C. SURROGATE KEY
A Surrogate Key’s only purpose is to be a
unique identifier in a database, for example,
B. SECONDARY KEY
incremental key, GUID, etc. It has
• Secondary Key is the key that has not
Surrogate Key has no actual meaning and is
been selected to be the primary key.
used to represent existence. It has an
However, it is considered a candidate
existence only for data analysis.
key for the primary key.
Let us see an example −
• Therefore, a candidate key not selected
as a primary key is called secondary
key. Candidate key is an attribute or set
of attributes that you can consider as a
Primary key.
D. FOREIGN KEY Language). Relational algebra operations are
used to perform various tasks on sets of
• A Foreign Key creates a link between
relations (tables) in a database.
tables. It references the primary key in
another table and links it. Fundamental Operators
• For example, the DeptID in the 1. Selection(σ)- The selection operation is
Employee table is a foreign key − used to retrieve rows from a table that satisfy
a given condition.
Let us see an example −
Notation − σp(r)
Where σ stands for selection predicate and r
stands for relation. p is prepositional logic
formula which may use connectors like and,
or, and not. These terms may use relational
E. UNIQUE KEY operators like − =, ≠, ≥, < , >, ≤.
• Many users consider Primary Key as
Unique Key, since both uniquely
identify a table, but Unique Key is
different from Primary Key. Unique
Key accepts null values and Primary
Key cannot have null.
• Let us compare Primary Key and
Unique Key and understand its concept
−
2. Projection(π)
The projection operation is used to retrieve
specific columns from a table.
5. Intersection (∩)
The intersection operation returns a relation b. Natural Join - If we join R1 and R2 on
containing only the rows that are common to equal condition then it is called natural join
both input relations. It displays the common or equi join. Generally, join is referred to as
values in R1 & R2. It is denoted by ∩. natural join.
6. Join (⨝)
The join operation combines rows from two Relationships within the Relational
relations based on a common attribute. Database
Different types of joins include Theta Join, In a relational database, relationships are
Natural Join, Outer Join (Left Outer Join, established between tables to organize and
Right Outer Join, Full Outer Join) link related data. The relationships are
defined using keys, which are columns or sets
of columns that uniquely identify records A junction (or associative) table is used to
within a table. There are several types of manage this relationship.
relationships in a relational database:
Example: A "Student" table and a "Course"
Several types of relationships in a relational table connected by a "StudentCourse"
database: junction table.
1. Primary Key - Foreign Key These relationships help maintain data
Relationship: integrity, avoid data redundancy, and provide
a way to navigate and retrieve related
•This is the most common type of
information across multiple tables in a
relationship.
relational database.
The primary key in one table is referenced as
a foreign key in another table.
Lesson 4
It establishes a link between the two tables.
ENTITY RELATIONSHIP DIAGRAM
Example: Connecting an "Employee" table
with a "Department" table using the An Entity Relationship Diagram is made up
DepartmentID as a foreign key in the of many different components:
Employee table.
• Entity
2. One-to-One Relationship: • Relationship
Each record in the first table is related to one • Attribute
and only one record in the second table, Symbols and notations
and vice versa.
Chen
Example: Storing detailed information about
employees in a separate table from their basic Example of the symbols used with the Chen
information. notation style.
3. One-to-Many Relationship:
Each record in the first table can be related to
multiple records in the second table, but each
record in the second table is related to only
one record in the first table.
Example: One department has many
employees, but each employee belongs to
only one department.
4. Many-to-Many Relationship:
Each record in the first table can be related to Crow’s Foot
multiple records in the second table, and vice
versa.
Barker
Logical
Bachman
Physical