- High level data model:
- This type of model omits a lot of details.
It is typically used to let non-technical users to talk about database issues. An example of a high conceptual model is a model that describes "how users perceive the data" (how they use it)
- Implementation data model:
- This type of model
has more details on
what data is
stored in the database and
how (what files) the
data is stored
- The model contains enough detail to let user see the structure of the data and how the data is stored inside the database.
This type of model is used in database systems to describe the structure of the data
The model is abstract enough to be used to let non-technical users to talk about (= discuss) the database details and also provides enough details to know how the data can stored.
- This type of model
has more details on
what data is
stored in the database and
how (what files) the
data is stored
- Low level data model:
- This type of
model contains
minute (very precise) detail
on how the
data is
store
- Examples of some of the
details:
- File names
- Indices on the files
- Statistics on the data in files (e.g.: number of records, record length, etc)
- Which disk will be used to store the data
This type of model is used in database system to describe how the data is stored inside the computer
- This type of
model contains
minute (very precise) detail
on how the
data is
store
- Fact:
- A database system
is used by users from
very diverse backgrounds:
- from Computer Science nerds
- to Managers who have no idea what a "byte" is....
- A database system
is used by users from
very diverse backgrounds:
- As such, a
database system
will use different levels of
models
to
describe the
data
stored inside the
database
- Jargon:
database schema
- Database schema =
a description
of a database
(It's actually a description of the data stored inside a database)
A database schema is in fact the same thing as a database model...
- Database schema =
a description
of a database
- A database schema
is specified (= constructed)
during the
design phase
of a database
- A Database schema will rarely be updated (changed)
A database schema is only updated when the information need of a company/university/etc changes
- To accommodate
many different groups
of users of a
database,
the database schema
(= description of the database)
uses the
3 different types of schemas :
- A high level description (schema)
called
external schema:
- Used by non-experts (= users of the DBMS) to specify their needs
- A intermediate level description
(schema)
called
conceptual schema
- Used by the designer of the database to specify how the data will be stored inside the DBMS
- A low level description (schema)
called
internal schema
- Used by the database administrator
to specify how the
files used in the
conceptual schema will
be stored
How a file is stored will determine many properties:
- How fast can the data in the file be accessed ?
- How much space will be needed to store the file.
- And so on.
- Used by the database administrator
to specify how the
files used in the
conceptual schema will
be stored
- A high level description (schema)
called
external schema:
- External schema:
- An external schema describes
how
a class of users
views/perceives
their
data
- Different users will have different needs and have different
perceptions of the data
E.g., a CEO and a clerk of the same company will have different needs and perceptions of the same company
- Different users will have different needs and have different
perceptions of the data
- There are
multiple
external schemas
in one
database schema (or database description)
- The
External schema
is an example of
high level data models
- An external schema describes
how
a class of users
views/perceives
their
data
- Conceptual schema:
- A Conceptual schema
describes the
structure
of the
data files in a
database
without giving
details on
how
the data files
are stored
- A conceptual schema must
encompass
all the
external schemas
in the database
In other words:
- You must be able to obtain each external view through a mapping/merging operation (with a computer program) using the conceptual schema.
- A conceptual schema must
encompass
all the
external schemas
in the database
- There is
only one
Conceptual schema for each database.
- The Conceptual schema is an example of the implementation data model
Note:
- The Conceptual schema is
the most important
schema (data model)
in a database system
- The Conceptual schema must be designed with great care !!!
We will learn more about how to design the Conceptual schema in the next 2 chapters
Also, you do NOT allow an average Joe to get his hand on the conceptual schema
The Database Administrator (DBA) is usually responsible for managing the conceptual schema.
- A Conceptual schema
describes the
structure
of the
data files in a
database
without giving
details on
how
the data files
are stored
- Physical schema:
- The Physical schema
describes
every detail on
how
the data files
of the database
are
stored inside the computer
.
Including:
- Format of each file
- Identifying fields (e.g., SSN, Student number) of each file
- Indexes (used to access record with specific values quickly)
- And so on.
- There is
only one
Physical schema for each database.
- The Physical schema is an example of the low level model
- The Physical schema
describes
every detail on
how
the data files
of the database
are
stored inside the computer
.
- Suppose a university (like Emory) wants to
maintain information on Students.
- The information system must support various applications:
- Student Population
- Health records
- Grade Reports
- Etc, etc...
- The following figure shows the 3 levels of data decription
of the data stored in such a database system:
- Each external schema
describe what information
ONE particular type of users
require
- The conceptual schema
list (describe) the
fields in each data file
- The internal schema describe the structure of each data file and how BEST to access each data file
- Each external schema
describe what information
ONE particular type of users
require
- The
physical data independence
technique
enables us to write
software to
"map":
- A Conceptual schema to a Physical schema
- In other words:
-
The database software uses meta data
that describes the
Conceptual schema
to access
the file structures of the
Physical schema
- The file(s) containing the meta data is called the
Database Catelog
- The
logical data independence
technique enables us to write
software to
"map":
- Any External schema onto the Conceptual schema
- In other words:
-
The database software allows users to supply computer programs
to manipulate the data files defined in the
Conceptual schema
to output
a file structure defined by a particular
External schema
- Each programs defined by the users
used to map the conceptual schema (description) to
an external schema is called
view definition
We don't have enough background to fully understand view definitions yet; when we get to SQL, we will learn more about view.
For the curious, here is the class notes on views: click here