Chapter 2: ER-Diagrams. Content: Learn how to draw ER diagrams Useful to model a database

Size: px

Start display at page:

Download "Chapter 2: ER-Diagrams. Content: Learn how to draw ER diagrams Useful to model a database"

April Adams
5 years ago
Views:

1 Chapter 2: ER-Diagrams Content: Learn how to draw ER diagrams Useful to model a database 1

2 Database Design DBS can take care automatically of many things but the user has to specify Requirements of the application Characteristics of the data Two important concepts during DBS design: Data Model: How to describe the data? Data Schema: Concrete description of the data (using the chosen data model) 2

3 Data modeling Excerpt of the Real World Conceptual Schema (E/R- or UML-Schema) Manual/intellectual Modeling Semi-automatic Transformation Relational Schema XML Schema Network Schema Object-orientierted Schema 3

4 Modeling a small example application: E/R Students Lectures Real World: University Conceptual Modeling requires StudNr Name Students attend Lectures LectureNr Title 4

5 Logical Data Models Network Model Hierarchical Model Relational Data Model XML Model Object-orientierted Data Model Object-relational Schema Deductive Data Model * [Michael Stonebraker: What Goes Around Comes Around] 5

6 Relational Data Model Students StudNr Name Fichte Jonas attend StudNr Lecture Nr Lecture Nr Lectures Title Grundzüge Glaube und Wissen... Select Name From Students, attend, Lectures Where Students.StudNr = attend.studnr and attend.lecturenr = Lectures.LectureNr and Lectures.Title = Grundzüge ; 6

7 Modeling a small example application: UML Students Real World: University Lectures Students +StudNr : int +Name : String +Semester : int +GPA() : float +SumWeeklyHours() : short +Attendee 1..* attend * +Successor * Lectures +LectureNr : int +Title : String +WeeklyHours : int requires * +NumberAttendees() : int +FailureRate() : float 7

8 Database Design Database Abstraction Layers 1. Conceptual Design 2. Logical Design 3. Physical Database Design 8

9 Phases of Database Design Information Requirements Att1 Relation Att2 Processing Requirements Requirements Engineering Scope Statement Conceptual Modeling Conceptual ER Schema Design Logical Modeling Logical Schema Logical Design Physical Modeling Physical Database Design DBMS- Characteristics Hardware/OS- Characteristics 9

10 Software Development and Ability to Communicate 10

11 Schema Design Approach in principle: Information Requirements Semantical Data Modeling Coarse Grain Data Modeling Semantical Fine Grain Data Analysis Modeling Logical Data Modeling Database Installation / Tuning Time - Interview - Brainstorming - Document s Analy. - - ERM - Hierarchical - UML - Conceptual Schema Design - network - relational - object-oriented -. Logical Schema Design - IMS - UDS - DB2 - Ozone - Physical Schema Design 11

12 Requirements Engineering Entity description Relation description Process description 12

13 Entity Description University Employees -Quantity: Attributes vempnumber Type: Integer Domain: Defined: 100% Identifying: yes Example: 007 vsalary Type: decimal Length: (7,2) Unit: Euro per month Defined: 10% Identifying: no vlevel Type: String Length: 2 Defined: 100% Identifying: no Example: W2 13

14 Relation Description: exam Involved Objects: - Professor as Tester - Student as Testee - Lecture as Test Subject Attributes of the Relation: - Date - Time - Grade Quantity: per year 14

15 Process Description : Issue a Certificate - Frequency: semiannually - Required Data * Tests * Examination Rules * Student s Records *... - Priority: high - Data Volume to be processed * 500 Students * 3000 Tests * 10 Versions of Examination Rules 15

16 Creating a Specification The actual analysis is an iterative process: Customer tells developer his/her needs Developer notes everything down (s/he understood) in his/her language" and translates it into the " language" of the customer This is shown to the customer who does not agree with everything Change requests are agreed on Back to step 2 16

17 Phases of Database Design Information Requirements Att1 Relation Att2 Processing Requirements Requirements Engineering Scope Statement Conceptual Modeling Conceptual ER Schema Design Logical Modeling Logical Schema Logical Design Physical Modeling Physical Database Design DBMS- Characteristics Hardware/OS- Characteristics 17

18 Conceptual Design The ideal design (the ideal specification) is unique complete comprehensible (for all participants) nonredundant... and not reachable in reality 18

19 Entity/Relationship-Modeling Entity Relationship Attribute (property) Key (identification) StudNr Name Semester Students attend Attendee Lectures Role LectureNr Title Weekly hours 19

20 University Schema StudNr Name Students prerequisite attend require Lectures followup LectureNr Weekly Hours Semester Title Grade test give PersNr Name Area Assistants work-for Professors PersNr Name Level Room 20

21 21

22 Functionalities E 1 R E 2 R Í E 1 x E 2 1:1 E 1 E 2 1:N N:1 N:M 22

23 Relationship 1:1 Relationship 1:1 E 1 1 R 1 E 2 e 1 out of E 1 takes part in 1 relation of type R e 2 out of E 2 takes part in 1 relation of type R Example: Car 1 has 1 License Plate one car has one license plate one license plate belongs to one car 23

24 Relationship 1:N Relationship 1:N E 1 1 R N E 2 e 1 out of E 1 takes part in N relations of type R e 2 out of E 2 takes part in 1 relation of type R Example: Mentors 1 advise N Students one mentor advises several students one student is advised by one mentor 24

25 Relationship N:M Relationship N:M E 1 N R M E 2 e 1 out of E 1 takes part in M relations of type R e 2 out of E 2 takes part in N relation of type R Example: Actors N stars M Movies one actor stars in several movies one movie has several actors 25

26 Recursive Relationship 1:N Relationship 1:N E 1 e 1 out of E 1 takes part in role A in N relations of type R e 1 out of E 1 takes part in role B in 1 relation of type R Example: Persons role A role B mother child rel-ship one person is mother of several persons (children) one person is child of one person (mother) 1 N 1 N R 26

27 Functionalities in n-ary Relationships E 1 P N M E n R E 2 1 E k R : E 1 x... x E k-1 x E k+1 x... x E n E k 27

28 Example Seminar Students N supervise 1 1 Professors Topics Grade supervise : Professors x Students Topics supervise : Topics x Students Professors 28

29 Thereby induced Consistency Constraints 1. Students may work on only one topic with the same professor (to cover a broad spectrum) 2. Students may work on the same topic only once thus they may not work on the same topic again with another professor 3. Professors can reuse the same topic i.e. give the same topic to different students 4. One topic can be given by different professors but to different students 29

30 Occurrence of the Relationship supervise p 1 Students b 1 p 2 Professors s 1 b 2 p 3 s 2 b 3 p 4 s 3 b 4 s 4 Dashed lines represent illegal occurrences b 5 b 6 t 1 t 2 t 3 t 4 Topics 30

31 One more Example 3-ary relationship: Patients perform N 1 Checkup 1 Experts One checkup is performed by one expert with several patients One Patient gets only one checkup from one expert One checkup is performed at one patient only by one expert 31

32 University Schema StudNr Name Semester Students N N prerequisite attend M require followup N M M Lectures M N LectureNr Weekly hours Title Grade test give PersNr Name Area Assistants N work-for Professors PersNr Name Level Room 32

33 (min, max)-notation E 1 E n (min n, max n ) R (min 1 max 1 ) (min 2, max 2 ) E 2 (min i, max i ) E i R Í E 1 x... x E i x... x E n For every e i Î E i there are at least min i tuples (..., e i,...) Î R and at most max i tuples (..., e i,...)î R 33

34 Example (min, max) Mentors advise [0, 20] [1, 1] Students one mentor advises up to 20 students one student is advised by exactly one mentor 34

35 Excercise for next class Inform yourself about unary binary ternary relationships Discussion / new examples next class! 35

36 Weak Entities Height BldNr RoomNr 1 N Buildings located_in Rooms Size Relationship between "strong" and "weak " type is 1:N (or 1:1 in rare cases) - why not N:M? The existence of a room depends on the existence of the associated building RoomNr is unique only within the building Key of Rooms is: RoomNr and BldNr 36

37 Tests as weak entity type Students 1 write N Tests Grade TestId StudNr consist-of N N give LectureNr M M PersNr Lectures Professors Several professors design one test Several lectures are inquired in one test 37

38 Generalization Generalization / Specialization: G Is-a S S is a specialization of G Example: animal Is-a cat 38

39 Generalization University Name University-Members is-a Students Employees PersNr StudNr is-a Area Assistents Professors Level Room 39

40 Conclusion University schema with generalization and (min, max)- notation è Nextpage 40

41 StudNr Name Semester Area Students attend (0,*) (3,*) require (0,*) (0,*) prerequisite (0,*) (0,*) Lectures followup (1,1) Grade test give (0,*) (0,*) (1,1) (0,*) Assistants work-for Professors LectureNr Weekly hours Title Level Room PersNr Name is-a Employees

42 Aggregation Bikes Part-of Part-of Frames Wheels Part-of Part-of Part-of Part-of Pipes Handlebars Rims Spokes

43 Non-mot. Vehicles is-a Vehicles is-a Aggregation and Generalization Mot. Vehicles is-a Bikes Part-of Motorbikes Part-of Automobiles Frames Wheels Part-of Part-of Part-of Part-of Pipes Handlebars Rims Spokes

44 Where are we? Information Requirements Att1 Relation Att2 Processing Requirements Requirements Engineering Scope Statement Conceptual Modeling Conceptual ER Schema Design Logical Modeling Logical Schema Logical Design Physical Modeling Physical Database Design DBMS- Characteristics Hardware/OS- Characteristics

45 Min,max Notation and Functionalities Polygon N (3, *) squareid borders M (1, 2) Edge N (2, 2) edgeid begin/end M (2, *) Point x y z Min-max: A Polygon has at least 3 Edges. An Edge has 1 or 2 Polygons. 45

46 Design criteria Rules for classification of entities and attributes: Entities should contain descriptive information Multi valued attributes should be classified as entities Attribute should be assigned to that Entity which describes it most directly Redundant relationships should be avoided However, it always depends on the application

47 Example: Order As entity: Client 1 places N Order M for_a As relationship: Client M order N Product 1 Product As attribute: orderid Client M buy N Product

48 Exercise Sheet 2

49 Exercise Sheet 2 address #beds name Hospital 1 consists_of N Department 1 has M employs N Employee is_a N works salary id is_a N Shift 1 data from to M runs number number N Station 1 contains N Room 1 Nurse N supervises M N Doctor 1 has

50 Exercise Sheet 2 A M R 1 B N c

Data modeling. Excerpt of the Real World. Manual/intellectual Modeling. Conceptual Schema (E/R- or UML-Schema) Transformation

Data modeling. Excerpt of the Real World. Manual/intellectual Modeling. Conceptual Schema (E/R- or UML-Schema) Transformation Data modeling Excerpt of the Real World Conceptual Schema (E/R- or UML-Schema) Manual/intellectual ll l Modeling Semi-automatic Transformation Relational XML etwork Object-orientierted Sh Schema Sh Schema