Database Beginnings
CIS 121 – Computer Concepts II
Instructor: Ron Christensen
What is Data?
• Data are facts about things, places, events
• Facts about data are called metadata
• Data are building blocks for information
3/10/2016
CIS 121 - Portland Community
College
2
Common Data Applications
• Word Processors – Like fifth generation
language, very difficult to track
• Spreadsheets – Mostly two dimensional.
When items change, multiple spreadsheets
become problem some.
• Databases – More efficient, real-time, online.
3/10/2016
CIS 121 - Portland Community
College
3
DBMS –
Database Management System
• DBMS software…
–
–
–
–
–
3/10/2016
Manages metadata
Organizes data
Enforces security
Manages access to data and metadata
Enforces integrity
CIS 121 - Portland Community
College
4
Figure 1.2
Information System
Input
Output
Database
3/10/2016
CIS 121 - Portland Community
College
5
What is a Database (Black Box)?
• A computer accepts input, processes it
according to some rules, and makes output.
• Everything will be represented as data,
including the input and the output
• So what the DBMS is doing is taking the
data produced (the input) and converting it
into information (the output)
3/10/2016
CIS 121 - Portland Community
College
6
Relational DBMS
• Relational refers to the method of
organizing data, i.e. in tables
• Developed in the1970s
• Most common of several different types of
DBMS systems
– Hierarchical
– Network
– Object
3/10/2016
CIS 121 - Portland Community
College
7
Microsoft ACCESS
• Relational DBMS software for Windows
• Available as part of MS-Office
3/10/2016
CIS 121 - Portland Community
College
8
Type of Databases
• Production Database
– Transaction oriented
– Lots of well-defined read/write/update actions
• Decision Support Database
– Query oriented
– Complex queries
– Few updates or writes
3/10/2016
CIS 121 - Portland Community
College
9
A Bigger Picture…
• Databases are a critical part of information
systems
• Information systems help people solve
problems
• Your personal database and web page will
form the basis of a personal information
system
3/10/2016
CIS 121 - Portland Community
College
10
Figure 1.2
Information System
Software
Operating System
Application
Programs
People
DBMS
Write
and
Use
Database
Design
and
Manage
Hardware
3/10/2016
CIS 121 - Portland Community
College
11
Entities
• Entities are anything about which you want
to know something
– People, places, things, events, concepts
• Data are facts about entities
3/10/2016
CIS 121 - Portland Community
College
12
Entity Sets
• Entity sets are collections of related
entities. Entities are related by their
classification:
• student entities are related by the fact that they
are all students
• invoice entities are related by the fact that they
are all invoices
• car entities are related by the fact that they are
all cars
3/10/2016
CIS 121 - Portland Community
College
13
Entity Sets, cont.
•Entity sets are named as a broad definition of
the whole entity.
•Entity and Entity set names are singular.
•Entity and Entity set names are capitalized.
Examples:
An entity named HR contains HR entities.
An entity named EMPLOYEE contains employee records.
An entity named DEPARTMENT contains department
records.
3/10/2016
CIS 121 - Portland Community
College
14
Entity Sets, cont.
• Entity Sets can only contain related records
– a STUDENT entity set may not contain
INVOICE data
– a DEPARTMENT entity set may not contain
invoice data
– a PRODUCT entity set may not contain
employee data
…. And so on
3/10/2016
CIS 121 - Portland Community
College
15
Relational Databases
• Data is stored in TABLES (also called
relational sets)
• Tables are collections of attributes for
related records
• The reality is that lots of different things
get stuffed into a table for a variety of
reasons
• (Tables are another word for entity sets)
3/10/2016
CIS 121 - Portland Community
College
16
Database Tables
• A database table is used to store a record
(or tuple)
– An entity set is a collection of related records
• An record is anything you want to keep track
of, so an entity may be a person, place, thing,
event, etc.
– (Record is another word for Entity)
3/10/2016
CIS 121 - Portland Community
College
17
Database Table Components
• At the conceptual level, a database table may be
viewed as a matrix.
– Matrix rows are also known as tuples or records
• each row contains an record
– Matrix columns are also called fields or attributes.
• Each column (field) contains the record’s attribute values
3/10/2016
CIS 121 - Portland Community
College
18
Attributes
• Attributes are the characteristics that
describe records
– A STUDENT record may be described by
attributes that may include...
•
•
•
•
•
3/10/2016
social security number
name
address
date of birth
major
CIS 121 - Portland Community
College
19
Records, Attributes, Data
• Records are things about which you want
to know something, e.g. STUDENT
• Attributes describe something about the
entity, e.g. the name of the student
• Data are the values of the attributes, e.g.
DOE, JOHN
3/10/2016
CIS 121 - Portland Community
College
20
Null
• A null is an absence of value
• A null is NOT
– A blank
– A zero
• A null has several meanings in a database
– Value does not exist
– Value is not known
3/10/2016
CIS 121 - Portland Community
College
21
Attribute Names
• Attribute names are capitalized.
• For documentation reasons, attribute
names are composed of two parts:
– the first few characters reflect the entity they
help describe.
– subsequent characters are sufficiently
descriptive to identify the attribute.
3/10/2016
CIS 121 - Portland Community
College
22
Attribute Names, cont.
• Examples of attribute names:
EMP_LNAME = employee last name
STU_GPA = student grade point average
PROD_CODE = product code
CUST_LNAME = customer last name
INV_NUM = invoice number
3/10/2016
CIS 121 - Portland Community
College
23
Keys
• Primary Key (PK)
– an attribute (or combination of attributes) that
uniquely identifies each row (tuple) in a table.
• A PK composed of two or more attributes is known
as a composite PK.
• Foreign Key (FK)
– an attribute in one table whose values match the
PK values in a related table or whose “values” are
null.
• FKs are used to link (connect) related tables.
3/10/2016
CIS 121 - Portland Community
College
24
Data Integrity.
• Entity Integrity
– PK uniquely identifies each entity in a table
• PK may not include nulls
• Referential Integrity
– FK values in one table match the PK values in the
related table
• FK may not include nulls
3/10/2016
CIS 121 - Portland Community
College
25
A Conceptual View of a
Database Table
Tables are named. The table you
see here is the EMPLOYEE table.
Named attributes (fields)
Each column contains the values of an
attribute. The EMP_FNAME column only
contains employee first names;
the EMP_PHONE may only contain
employee phone numbers.
Each row represents
an record
A row is also called
a tuple.
The Primary Key (PK) is a unique record
identifier. If you know the PK value, you
will know all of its CIS
row’s
attribute values
3/10/2016
121 - Portland Community
College
Each row/column
intersection contains
only one of an entity’s
attribute values
26
Entity Integrity
EMPLOYEE table
A table exhibits entity integrity when all of its
Primary Key (PK) values uniquely identify each
table row (record.)
1. A PK cannot contain duplicate values
2. A PK cannot contains nulls
Note: A null indicates the absence of a value; it is not a blank. (You create
a null when you tap the ENTER key without first making an entry. A blank
is created when you tap the space bar and then tap the ENTER key.)
3/10/2016
CIS 121 - Portland Community
College
27
Foreign keys (FK)
and Referential Integrity
A Foreign Key (FK) is an attribute located in one table that “points to” a Primary
Key (PK) in a related table. The use of FKs allows you relate one table to another.
To maintain referential integrity, a foreign key (FK) must reference an existing
PK value in a related table it may be null.
INVOICE table
LINE table
3/10/2016
CIS 121 - Portland Community
College
28
Attribute types
1. Simple (atomic)
2. Composite
A. Single-valued
B. Multi-valued
Derived
3/10/2016
CIS 121 - Portland Community
College
29
Simple (Atomic) Attributes
• A simple (atomic) attribute cannot be
decomposed into meaningful components
• Examples:
– The attribute EMP_LNAME cannot be
decomposed, because you cannot subdivide
EMP_LNAME into a set of new attributes.
– The attribute PROD_PRICE cannot be
decomposed, because you cannot subdivide
PROD_PRICE into a new set of attributes.
3/10/2016
CIS 121 - Portland Community
College
30
Simple (Atomic) Attributes,
cont.
Simple attributes may be ….
• single-valued
or
• multi-valued
3/10/2016
CIS 121 - Portland Community
College
31
Simple (Atomic) Attributes,
cont.
• Single-valued simple attributes
• Example: an employee can have only one gender,
so EMP_GENDER is a single-valued attribute.
The attribute EMP_GENDER cannot be
decomposed, so it is a simple attribute.
• Multi-valued simple attributes
• Example: an employee can have many degrees,
so EMP_DEGREE is multi-valued. The attribute
EMP_DEGREE cannot be decomposed, so it is a
simple attribute.
3/10/2016
CIS 121 - Portland Community
College
32
Composite Attributes
•A composite attribute can be decomposed into
meaningful components
• Example: an employee’s address, shown as
123 East Main Street, Nashville, TN 32123
may be decomposed into
EMP_ADDRESS = 123 East Main Street
EMP_CITY = Nashville
EMP_STATE = TN
EMP_ZIP = 32123
3/10/2016
CIS 121 - Portland Community
College
33
Composite Attributes, cont.
A composite attribute may be ….
• single-valued
or
• multi-valued
3/10/2016
CIS 121 - Portland Community
College
34
Composite Attributes, cont.
• single-valued composite attributes
– Example: an employee can have only one date of
birth, so EMP_DOB is single-valued. But the
attribute EMP_DOB can be decomposed into year,
month, and day, so it is a composite attribute.
• multi-valued composite attributes
– Example: an employee can have more than one
address, so EMP_ADDRESS may be multi-valued.
The attribute EMP_ADDRESS can be decomposed
into street address, city, state, and ZIP code, so it is
a composite attribute.
3/10/2016
CIS 121 - Portland Community
College
35
Attribute Storage
• Each table row/column intersection contains a
single attribute value for a single entity.
– Ideally, attributes are simple and single-valued.
– Single-valued composite attributes are acceptable
• but composite attributes may make queries more complex
and may impose reporting limitations.
– Multi-valued attributes, either simple or composite,
may
• create structural problems
• make queries more complex and may impose reporting
limitations.
• May be necessary for a variety of reasons, but should be
avoided where possible
3/10/2016
CIS 121 - Portland Community
College
36
Multi-valued Attribute Storage
• Multi-valued attributes are sometimes
stored as
– strings
• this approach yields query complexity and reporting
limitations
– multiple attributes
• this approach yields structural problems
• Ideally, multi-valued attributes are handled
through the use of composite tables.
3/10/2016
CIS 121 - Portland Community
College
37
Multi-valued Attributes Stored
As Strings
Multi-valued
attribute
Poor practice: Makes it difficult to generate queries such as
“How many employees have earned BA or MBA degrees?”
3/10/2016
CIS 121 - Portland Community
College
38
Storing Multi-valued Attributes
In Separate Columns
Poor structure:
many nulls
table structure must be altered when additional degrees are earned
3/10/2016
CIS 121 - Portland Community
College
39
Composite Tables
• Multi-valued attributes can generate many
nulls, increasing uncertainty about the data
• One solution is to add another table, called
a composite table or bridge table
• It is often better to use more tables rather
than poor attributes in fewer tables
3/10/2016
CIS 121 - Portland Community
College
40
A Composite Table (EDUCATION) is Used
To Convert M:N Relationships To 1:M Relationships
Table name:
EMPLOYEE
Table name:
EDUCATION
Table name: DEGREE
3/10/2016
CIS 121 - Portland Community
College
41
Supertype/Subtype Relationships Are Used
To Eliminate or Control the Occurrence of Nulls
A table with many nulls
Supertype. (Table name: EMPLOYEE
Subtype. (Table name: PILOT)
3/10/2016
CIS 121 - Portland Community
College
42
Relationships
• Tables can be related to each other in a
variety of ways
–
–
–
–
1:1 Relationships
1:M Relationships
M:N Relationships
Recursive Relationships
• Tables are related by their PKs and FKs
3/10/2016
CIS 121 - Portland Community
College
43
1:1 Relationships
• Entity in one table is related to only one
entity in another table
• Example
– STUDENT : SOCIAL SECURITY NUMBER
Each student has only one Social Security Number, each Social
Security Number refers to only one student
3/10/2016
CIS 121 - Portland Community
College
44
1:M Relationships
• Entity in one table is related to many
entities in another table
• Example
– STUDENT : TUITION INVOICE
Each student may have several tuition invoices, each tuition
invoice is assigned to only one student
3/10/2016
CIS 121 - Portland Community
College
45
M:N Relationships
• An entity can occur more than once on
both sides of the relationship – causes
several different problems in databases
• Example:
– STUDENT : CLASS
Each student may register for many classes, each class may
contain many students
3/10/2016
CIS 121 - Portland Community
College
46
Some Issues with M:N
• May signal redundant data, resulting in
anomalies
– Modification anomaly
– Deletion anomaly
• Keep it simple: DO NOT IMPLEMENT
M:N RELATIONSHIPS. Use composite
tables instead
3/10/2016
CIS 121 - Portland Community
College
47
Recursive Relationships
• Entity is related to itself
• Examples
– TEAM : TEAM
Teams in an athletic league play each other
– CLASS : CLASS
Some classes are pre-requisites for others
3/10/2016
CIS 121 - Portland Community
College
48
Optional (vs. Mandatory)
Relationships
• Optional relationships occur when entities
in two tables may be related, but do not
need to be.
• Example (Optional Relationship):
– STUDENT : SOCIAL SECURITY NUMBER
A student is not required to have a social security number
• Example (Mandatory Relationship)
– STUDENT : CLASS
A student must be registered in a class to be a student
3/10/2016
CIS 121 - Portland Community
College
49
End of Lecture
3/10/2016
CIS 121 - Portland Community
College
50