Database design - relational data theory (ultra detailed)

？ problem —— What is a good database logic design ？

● Logical design of relational database ：
* How to construct a suitable data schema for a specific problem , That is, several relationships should be constructed , What are the attributes of each relationship
eg：

？ Is this design a good design ？

Observe an instance corresponding to this table （ At some point student An actual data situation corresponding to the pattern ）： If there are several students , They are all for ’ Department of Computer Science ’, The dean of the Department is ’ Zhang Ming ’, Optional ’C1’ Course , Get their own grades
*

* Relationship model STUDENT（Sno,Sdept,Mname,Cno,Grade） Problems in ：*
1、 Too much data redundancy , Waste storage space

Such as ： The name of the Dean reappears , The number of repetitions is the same as the number of occurrences of all course scores of all students in the Department .

2、 Update exception （Update Anomalies）
data redundancy , When updating data , Maintaining data integrity is costly
If a department changes its Dean , The system must modify every tuple related to the students of the Department .

3、 Insertion exception （Insertion Anomalies）, The data to be inserted cannot be inserted

If a new software engineering department is established , No admissions yet , We can't store the information of the Department and its Dean in the database .

4、 Delete exception （Deletetion Anomalies）, Data that should not be deleted has also been deleted

If all the students in a department graduate , While deleting the student information of the Department , Lose the information about the Department and its Dean .

？ What is a good model ？

* A good design pattern will not cause insertion exceptions 、 Delete exception 、 Update exception 、 Data redundancy should be minimized

The cause of the problem ：
Due to some in the pattern Data dependency Caused by the .
Data dependency :
◆ It is the relationship between data reflected by the equality of values between attributes in a relationship
◆ It is an abstraction of the interrelationship between real-world attributes
◆ Is the inherent nature of data
◆ Is the embodiment of semantics

The impact of data dependency on relational patterns ：
Inappropriate data dependencies , Cause insertion exception 、 Delete exception 、 Update exceptions and data redundancy problems

Use a graph to represent functional dependencies

resolvent ： Separate the first mock exam. 3 A relationship model
S（ Sno, Sdept, Sno→Sdept);
SC ( Sno, Cno, Grade, (Sno, Cno)→Grade);
DEPT (Sdept, Mname, Sdept→Mname)
These three modes will not cause insertion exceptions 、 Delete and other exceptions ; Data redundancy is controlled .

How to solve the problems in the relationship pattern ：
Normalization theory —— Identify inappropriate data dependencies in the relational schema , Eliminate them , Insertion exceptions can be resolved to varying degrees 、 Delete exception 、 Update exceptions and data redundancy problems .

Function dependency

How to determine functional dependencies ？
◆ Functional dependency is the concept of semantic category . Can only Determine the functional dependency according to the semantics of the data .
● Such as Sname→Sno Functional dependencies only exist in “ Students are not allowed to have duplicate names ” Under the conditions of .

Trivial and nontrivial functional dependencies
●X→Y, Y⊈X, said X→Y yes Nontrivial functional dependencies .
●X→Y, Y⊆X, said X→Y yes Trivial functional dependencies .

Complete functional dependency and partial functional dependency

The transfer function depends on

normal form

◆ A paradigm is a collection of relational patterns that conform to a certain level

◆ The relationship in relational database must meet certain requirements .
Different paradigms meet different requirements

● Types of paradigms ：
* First normal form （1NF)********（ Relationship model R All properties of are basic data items that cannot be separated ）
* Second normal form （2NF)********（ Relationship model R⊆1NF, also Each non primary attribute is completely functionally dependent on R Code of , be R⊆2NF.
* Third normal form （3NF)********( Relationship model R⊆1NF, if R There is no such code in X、 Attribute group Y And non attributes Z（Y⊉Z）, bring X→Y,Y→Z,Y↛X, establish , said R<U,F>⊆3NF ◆ if R⊆3NF, be R Each non primary attribute of is neither partially functionally dependent on candidate codes nor transitive functionally dependent on candidate codes
)
*BC normal form （BCNF,Boyce and Codd Jointly proposed paradigm )******** It is also generally believed that BC Paradigm is the modified third paradigm , Sometimes called extended third paradigm , That is, in the relationship mode R<U,F> in , If every determinant contains a code , be R⊆BCNF.
◆BCNF The nature of relational patterns ：
① All non main properties are completely dependent on each code .
② All main attributes are also fully functional dependent on each code that does not contain it .
③ No attribute completely depends on any set of non code attributes .

* Fourth normal form （4NF)********
* The fifth paradigm （5NF********
● There are links between various paradigms （ Low level paradigm includes high-level paradigm ）：

Standardized summary

● A relational model as long as it Components are inseparable data items , It is a standardized relationship model , But this is just the most basic normalization .
● The relationship model with low standardization may not be able to describe the real world well , There may be insertion exceptions 、 Delete exception 、 Modify exception 、 Data redundancy and so on , The solution is to standardize it , Convert to high-level paradigm .
● A relational model of a lower paradigm , Through pattern decomposition, it can be transformed into a set of relational patterns of several higher-level paradigms , This process is called Normalization of relationship pattern .
● The normalization theory of relational database is a tool for database logic design .

◆ The basic steps of relational schema normalization ：