SUMMARY OF MODULE 8

1. A conceptual model of data used in an application is obtained by using an entity

relationship model (E-R model).

2. Entities specify distinct real world items in an application, relationships

represent dependencies between entities, and attributes specify properties of

entities and relationships.

3. Entity set describes items which are similar. Students, courses, teachers are

examples of entity sets.

4. Examples of relationship are students attend courses, Teachers teach courses.

“Attend” and “teach” are relationships.

5. E-R models have the facility to describe 1 to 1, 1 to many, many to many,

conditional and optional relationships.

6. E-R models assist in designing relational databases.

7. A relation is a table which depicts an entity set. Each column in the relation

corresponds to an attribute of the entity. Each row contains a member of the

entity set.

8. A relational database consists of a collection of relations relevant for a specified

application.

9. Relations in a database are normalized. Normalization is a procedure used to

transform a set of relations into another set which has some desirable properties.

10. Normalization ensures that data in the database are not unnecessarily duplicated.

It also ensures that addition and deletion of entity rows (or tuples) or change of
 individual attribute values do not lead to accidental loss of data or errors in

database.

11. Normalization is performed by first converting a relation so that all entities are

simple. This results in a relation said to be in First Normal Form (INF).

12. Successive normalization steps are carried out to meet the objectives stated in

point 10.

13. The successive normal forms are called 2NF, 3NF, BCNF, 4NF and 5NF.

14. A summary of how normalization steps are carried out is shown in the table

below.

Summary of Normalization Steps

Input Relation Transformation Output relation

All relations Eliminate variable length INF

records. Remove
multiattribute lines in table
INF relation Remove dependency of 2NF
non-key attributes on part of
a multiattribute key
2NF Remove dependency of 3NF
non-key attributes on other
non-key attributes
3NF Remove dependency of an BCNF
attribute of a multiattribute
key on an attribute of
another (overlapping)
multiattribute key
BCNF Remove more than one 4NF
independent multivalued
dependency from relation
by splitting relation
4NF Add one relation relating 5NF
attributes with multivalued
dependency to the two
relations with multivalued
dependency
15. Traditionally, application programs were written after designing data files

appropriate to the application.

16. As organizations require many application programs and these applications need

some common data, it is found that the same data is stored in many files, thereby

wasting storage space.

17. Waste of storage space is not the only problem. The same data (such as an

address of an individual) may be in different forms in different files, leading to

confusion.

18. As data is scattered in many files, it is difficult to correlate them and use them for

various applications.

19. When requirements change, the application programs and the entire data

organization may have to be changed. Such a change is time consuming and

expensive.

20. Instead of designing independent files for different applications, it is better to

design a common database which can be used for all current applications and

applications expected to be implemented in the future.

21. A database constitutes the data resource of an organization. It incorporates not

only data but also relationships between data items. It is modelled not for specific

current applications, but all current and potential applications.

22. A Database Management System (DBMS) is a set of procedures that manage a

database and provide application programs access to the database in a form

requested by the applications programs.

23. A good DBMS allows sharing of data among applications. Sharing of data is

possible if application programs can be written without having a knowledge of

data representation, location or access. In other words, application programs

should be free to have their own view of data by DBMS.

24. Other major objectives of DBMS are to minimize storage of redundant data,

maintain consistency of data values, ensure validity of stored data, protect data,

prevent unauthorized access, control insertion, deletion and alteration of data and

ensure fast access to data required by application programs.

25. To use a DBMS, a conceptual model of data needed for current and future

applications is designed. E-R modelling is useful for this. The conceptual model

is next converted to a logical model appropriate for the DBMS. For RDBMS,

normalised relations constitute the logical model.

26. The physical data model is then designed to ensure good performance. User's

view of data for application is primarily the logical view.

 QUESTION BANK - MODULE 8

8.1 What is an entity? Give examples of entities.

8.2 What is a relationship? In what way is it different from an entity?

8.3 In what way is an attribute different from an entity?

8.4 What do you understand by a 1 to 1, 1 to many, many to many relationships?

Give one example for each.

8.5 What is the difference between a relation and a relationship?

8.6 What is normalization of a relation?

8.7 Why should relations be normalized?

8.8 What is a functional dependency?

8.9 What is a key attribute in a relation?

8.10 What is the difference between a 2NF and a 3NF relation?

8.11 When is BCNF required?

8.12 When are 4NF and 5NF required?

8.13 What is the difference between 4NF and 5NF?

8.14 Develop E-R diagram for the following:

Customer withdraws money from his account

Students write examinations.

Students attend classes

Professors write books

Driver drives a car

8.15 Draw an E-R diagram showing the cardinality for the following:

(i) A bill is sent to a customer. A customer can receive many bills.

(ii) A clerk works in a bank. The bank has many clerks

(iii) A part is used in many products and a product uses many parts.

(iv) Students apply for seats in colleges. Each student can atmost get one

seat. A college has many seats. A student can send many applications.

(v) A car is owned by a person. The person can own many cars.

8.16 For Exercise 8.14, obtain relations for each entity. Normalize the relations.

8.17 For the following word statement, obtain E-R diagram and relations. Use any

reasonable assumptions. "A machine shop produces many parts which it takes

on contract. It employs many machinists who operate any of the machines. A

part needs working on only one machine. A record is kept on the quantity of

material needed for producing each part. The production of each part is tracked

by giving a job number, start time and end time and machinist identifications".

8.18 For the problem on library, periodical management stated in Exercise 4.23

(Module 4), obtain E-R diagrams and relations. Make any reasonable

assumptions and state the assumptions.

8.19 For the problem statement of Exercise 5.15 (Module 5) obtain the E-R diagram

and relations for the problem.

8.20 For the problem statement of Exercise 5.16 (Module 5), obtain the E-R diagram

and a set of relations by using the diagram

8.21 For the problem statement of Exercise 5.18(Module 5), obtain the E-R diagrams
 and relations .

8.22 What are the advantages and disadvantages of systems using separate data files?

8.23 What do you understand by the term data integrity?

8.24 If redundant data is stored can it lead to data integrity problem?

8.25 Student's records in a University are kept by various sections: Hostel, Health

Centre, Academic Office, major departments, Accounts Section and Library. If

each of these sections maintains its own file-based system for processing, what

problems do you foresee? Give examples.

8.26 Define a database of an organization.

8.27 What is the difference between a database and a Database Management System

(DBMS)?

8.28 What are the basic objectives in evolving a database for an organization?

8.29 What do you understand by the term data independence?

8.30 What advantages are available to a programmer from data independence?

8.31 Draw an analogue between the advantages of high level language programming

and data independence of application programs.

8.32 How can data integrity be maintained in a database?

8.33 Distinguish between issues of privacy and security in a database.

8.34 What is the role of E-R diagrams in database design?

8.35 What is the difference between a conceptual model and a logical model of a

database?

8.36 What is an internal model of a DBMS?

8.37 What data models are used by application programs in a database oriented
 system?

8.38 Why is a DBMS divided into three layers, namely, conceptual model, logical

model, and internal model?

8.39 How is data independence of application programs ensured in a DBMS?

8.40 What is RDBMS?

8.41 What are the responsibilities of a Database Administrator (DBA) in an

organization?

8.42 What are the ideal features of a database system?

8.43 Are DBMS relevant to Personal Computers (PCs)?

 Module 8

Mini Case Example 1

We will illustrate the method used to obtain an E-R diagram with an example. Usually

an imprecise statement of the needs of an organization is given by the management. We

now give a sample of such a statement. This is to be used to obtain an E-R diagram.

User's narrative of requirements. "Our company receives a number of items from

many vendors and they are received at the receiving office. As we receive over 1000

items a day it is now virtually impossible for the receiving office to check whether the

vendor has supplied items against an order, or sent a wrong item and inform the purchase

office. We are also not able to find out if there are excesses or deficiencies in delivery

and whether the vendor adhered to the delivery schedule as per the order. The items

received at the receiving office are sent for physical inspection. The physical inspection

consists of checking whether the quantities stated in the delivery note agree with the

physical count, whether the item is the correct one ordered, and a check on the quality of

item. We would like to keep a record of rejections due to bad quality, incorrect items,

excess/deficient supply etc., determined during inspection. This will enable us to assess

vendors' reliability to guide us in placing orders in the future, besides keeping track of

supplies. Items cleared by the inspection office are taken into the inventory by the stores

office which keeps a ledger of items stocked and quantity available of each item.

Customers send requisitions to the stores. The stores fulfill the requests based on

availability and update the ledger. Currently we are not able to meet some of our
customers' requests. We would like to incorporate automatic reordering by our purchase

office if the inventory level of an ite

We would also like to keep track of unfulfilled requests and meet them when items

reach the store. Currently we are not able to pay our vendors promptly due to delays in

payment order reaching our accounts office. We would like to rectify this. We would

also like to bill our customers promptly and keep track of customers' payments".

The E-R diagram for the problem is:

orders

Placed Placed
for with

items vendors

Supplies

The relations are:

ORDER (order no.., order date)

ORDER PLACED FOR (order no., item code, qty. ordered, delivery time allowed)

ORDER PLACEDWITH (order no.. vendor code, item code)

VENDOR (vendor code, vendor name, vendor address)

ITEM (item code, item name, price/unit)

SUPPLIES (vendor code, item code, order no., qty.supplied, date of supply).

The key attribute(s) are in bold letter(s) in each relation.

Let us examine whether the relations are in normal form. ORDERS and ORDER

PLACED FOR are simple relations. In the relation ORDER PLACED WITH, the

key is the composite attributes order no., vendor code and item code. However, the

entity item code is not needed in the ORDER PLACED WITH relation. Given an

order no., all the items supplied against this order can be found from ORDER

PLACED FOR relation. The vendor with whom the order has been placed can be

found from the ORDER PLACED WITH relation given below as each order no., has

only one vendor.

ORDER PLACED WITH (order no., vendor code)

The two relations ORDER PLACED WITH and ORDER PLACED FOR have

composite keys. The non-key fields are not related to one another. In a key more

than one attribute, the individual attributes are not functionally dependent. Thus these

two relations are in normalized form and do not need any further change. The

relations VENDOR and ITEM are simple and are in normalized form. The relation

SUPPLIES is, however, not normalized. vendor code and order no., are

functionally dependent. There is a multivalued dependency between vendor code and

item code as a vendor can supply many items. We thus split the relations into two

relations.

ACTUAL VENDOR SUPPLY (order no., item code, qty.supplied, date of supply)
 VENDOR SUPPLY CAPABILITY (vendor code, item code)

Observe that the relation VENDOR SUPPLY CAPABILITY will have a (vendor

code, item code) table without a vendor having supplied any item. The relation

ORDER PLACED WITH will have a tuple only when a vendor actually supplies an

item.

Minicase Example 2

Let a database contain the following: Teacher code, Teacher’s name, Teacher’s

address, rank, department, courses taught by the teacher, course name, credits for

course, no.of students in the class, course taught in semester no., student no., name,

dept., year, and courses taken in semesters no. The following information is given on

dependencies.

• A teacher may teach more than one course in a semester

• A teacher is affiliated to only one department

• A student may take many courses in a semester

• The same course may have more than one section and different sections will

be taught by different teachers.

• A course may be taught in more than one semester

An entity relationship diagram for this problem is given below. The relations

corresponding to the E-R diagram are:

 Teacher Studnet

M K

Teaches Attends

N P
Courses

An E-R diagram for teacher database

TEACHER (Teacher code, teacher’s name, teacher’s address, rank, dept.)

TEACHES COURSES (Teacher code, course no., semester no., no.of students, section

no)

COURSE (course no., semester no., course name, credits)

STUDENT (student no., student’s name, dept., year)

STUDENT –COURSES (student’s no., course no., semester no.)

TEACHER relation has only one key. All non-key attributes are functionally dependent

only on the key. There is no functional dependency among non-key attributes. Thus the

relation is normalized in 3NF (No higher NFs are applicable).

STUDENT relation is also, similarly, in 3NF. In the COURSE relation, course name

could also be a key. The relation is in 3NF and no further normalization is required. The

relations TEACHES COURSES and STUDENT-COURSES have multiattribute keys, but

the relations themselves are in normal form. The only point which is not clear, from

these relations, is the relation between teacher and student. This has been missed in the

E-R diagram. The relationship is between the teacher, courses taught and students. In

other words, we should be able to answer the question “which teacher is teaching course

no.X to student no. Y in Semester 2?”. Let us add a relation.

TEACHER-STUDENT (Teacher code, student no., course no.)

In this relation Teacher code and course no. have a multivalued dependency. Similarly,

Teacher code and student no. as well as student no. and course no. have multivalued

dependency. However, TEACHES COURSES (Teacher code, course no.,semester

no.., no.of students, section no.) and STUDENT COURSES (student no., course no.,

semester no) relations are already in the database. Thus the relation TEACHER-

STUDENT as it is specified above is sufficient to give the idea that student Y takes

course X from Teacher Z.

References

1. Most of the material in this module has been adapted from the book “Analysis and

Design of Information Systems”, 2nd Edition, by V.Rajaraman. Chapter 10

Logical Data Base Design, pp.130 to 153. Chapter 12 , Data Base Management

Systems, pp. 171 –1719.

2. Good treatment of E-R Modelling and Normalization may be found in the book

“Introduction to Systems analysis and Design”, by I.T.Hawryszkiewycz, Prentice

Hall of India, 1989.

3. For those interested in detailed study of Data Base Design and DBMS, the

following books are standard texts:

1, R.Elmasri and S.B.Navathe, Fundamentals of Data Base Systems, 4th Edition,

Pearson Education Asia, New Delhi, 2004.

2. T.Connolly and C.Begg, Dta Base Systems, 3rd Edition, Pearson Education

Asia, New Delhi, 2003.