1 Chapter 9 Database Management Systems Accounting Information Systems, 5 th edition James A. Hall.

1

Chapter 9

Database Management Systems

Accounting Information Systems, 5th editionJames A. Hall

Flat-File Versus Database Environments

Computer processing involves two components: data and instructions (programs).

Conceptually, there are two methods for designing interface between programs and data: file-oriented (flat-file) processing: A specific data file

was created for each application [legacy systems] data-oriented processing: Create a single data repository

to support numerous applications [current systems] Disadvantages of file-oriented processing include redundant

data and programs and varying formats for storing the redundant data.

File-Oriented (Flat-File) Environment

Program 1

Program 2

Program 3

A,B,C

X,B,Y

L,B,M

User 2 (A/R)Transactions

User 1(Sales)Transactions

User 3 (Payroll)Transactions

Data

Where B is a customer record. Assume B’s address changes. Each userthat works with the B record must modify the address, increasing chance for error.

Program 1

Program 2

Program 3

User 2 (A/R)Transactions

User 1 (Sales)Transactions

User 3 (Payroll)Transactions

Database

DBMS

A,B,C,X,Y,L,M

Database Approach

With the database approach, any User could change B’s address - it would be changed for all users. changes. [Assume all users have authority to change theaddress for this example.]

Data Redundancy & Flat-File Problems

Data Storage - creates excessive storage costs of paper documents and/or magnetic form

Data Updating - any changes or additions must be performed multiple times (see example of B’s address)

Currency of Information - potential problem of failing to update all affected files

Task-Data Dependency - user’s inability to obtain additional information as his or her needs change

6

Advantages of the Database Approach

Data sharing/centralized database resolves flat-file problems:

No data redundancy - Data stored only once, eliminating data redundancy and reducing storage costs.

Single update - Because data are in only one place, requires only single update procedure, reducing time/cost of keeping database current.

Current values - Change to database made by any (authorized) user yields current data values for all other (authorized) users.

Task-data independence - As users’ information needs expand beyond immediate domain, new needs can be more easily satisfied than under flat-file approach.

7

Disadvantages of the Database Approach

Can be costly to implement additional hardware, software, storage, and

network resources are required – yet these costs continue to drop…

Can only run in certain operating environments may be unsuitable for some existing system

configurations (e.g., legacy systems)

8

Internal Controls and DBMS1

Purpose of DBMS is to provide controlled access to database.

DBMS is special software programmed to know which data elements each user is authorized to access and deny unauthorized requests of data.

1 DBMS = database management system

9

Query Language Query capability permits both end users and

professional programmers to access data (information) in database without writing conventional programs.

10

Three Steps in Designing a Database1. Prepare conceptual model

Identify entities Identify relationships between entities Prepare ER diagram (ERD)

2. Specify logical design Select logical database model (which will always be

relational nowadays) Transform conceptual data model using logical

database model

3. Implement physical design Physical structures Access methods

11

Phase 1Prepare the Conceptual Model

Draw an ERD to capture the process.

Example of a Relationship LinkingTwo Entities

CUSTOMERplaces ORDER

Name

Cust Number

Order Number Date

Example of 1:M relationship, using Crow’s Feet notation. We usually don’t cover the “0” relationship.

13

Phase 2Specify Logical Design

Create relational tables.

14

Logical Data Structures Objective is to develop structure efficiently so

data can be accessed quickly and easily. Four types of database structures are:

hierarchical (tree structure) network relational object-oriented

Concentrate on relational DBs. They are the systems being created at the current time.

15

The Relational Model Relational model portrays data in form of two

dimensional tables (looks like Excel worksheet): relation - database table attributes/fields (data elements) - columns tuples (records) - rows data - intersection of rows and columns

16

Properly Designed Relational Tables

No repeating values – Primary Key cannot repeat in a table.

Attribute values in any column must all be of same class. [can’t put text in a date field]

Each column (field name) in a table must be uniquely named. [can’t have 2 address fields both named Address – use Addr1 and Addr2]

Each row (record) in a table must be unique in at least one attribute (primary key)

17

Relational Model Data Linkages (>1 table)

No explicit pointers are present – data are viewed as collection of independent tables.

“Relations” formed by attribute/field that is common to both tables in relation. This field is a “foreign key.”

How to assign foreign keys: if 1:1 association, either of table’s primary keys may be foreign

key. if 1:M association, primary key of the ONE side is embedded

as foreign key in the MANY side. if M:M (M:N) association, create separate linking table.

18

Three Types of Anomalies(anomalies are found in Unnormalized tables)

Insertion Anomaly: New item cannot be added to table until at least one entity uses particular attribute item.

Deletion Anomaly: If attribute item used by only one entity is deleted, all information about that attribute item is lost.

Update Anomaly: Modification on attribute must be made in each of rows in which attribute appears.

Anomalies can be corrected by creating relational tables.

19

Relational Tables Various items of interest (customers,

inventory, sales) stored in SEPARATE tables in database.

Advantages: Removes all three anomalies Efficient use of space. Flexible. Users can form ad hoc relationships for

queries.

20

Normalization Process Process that breaks up large, complex tables

into smaller tables that meet two conditions: all nonkey attributes (fields) in table are

dependent on primary key (PK) all nonkey attributes (fields) are independent of

other nonkey attributes (fields) When unnormalized tables are split and

reduced to third normal form, they are linked together by foreign keys (secondary keys).

21

StudentID Lname Fname Major CRN Course

Instructor ID

80012 Garcia Maria ACCT 17256 ACCT3321 G001

80012 Garcia Maria ACCT 17535 ACCT3320 G003

80012 Garcia Maria ACCT 17004 MKT3311 G005

80014 Henry Hank MKT 17155 MKT3311 G005

80020 Hernandez Barbara CIS 17200 CIS2320 G010

80020 Hernandez Barbara CIS 17155 MKT3311 G005

80033 Johnson Kris ECON 16256 ACCT3321 S001

80033 Johnson Kris ECON 17005 ECON3315 S005

80033 Johnson Kris ECON 16500 CIS3545 M001

all nonkey attributes (fields) in table are dependent on primary key (PK)

This is an unnormalized table. Only Lname & Fname depend on the PK ofStudent ID. Other fields belong in another table(s).

Steps in Normalization

Unnormalized Table with repeating groups

First normalform (1NF)

Second normalform (2NF)

Third normalform (3NF)

1. Removerepeating

groups

2. Remove partial

dependencies

3. Removetransitive

dependencies

Does field depend on PK? If no, remove and put in another table.

Only concerns tables with composite PKs. See Fig. 9-38

When field depends on another non-key field in table. See Fig. 9-40

Accountants and Data Normalization

Update anomalies can generate conflicting and obsolete database values.

Insertion anomalies can result in unrecorded transactions and incomplete audit trails.

Deletion anomalies can cause loss of accounting records and destruction of audit trails.

Accountants should understand data normalization process and be able to determine whether database is properly normalized.

24

Phase 3Implement Physical Design

Decide about software and hardware.

25

Distributed Data Processing

26

President

VPMarketing

VP ComputerServices

VPOperations

VPFinance

SystemsDevelopment

DatabaseAdministration

DataProcessing

New SystemsDevelopment

SystemsMaintenance

DataControl

DataPreparation

ComputerOperations

DataLibrary

President

VPMarketing

VPFinance

VPOperations

IPU IPU IPU IPU IPU IPU

VPAdministration

Treasurer Controller ManagerPlant X

ManagerPlant Y

CENTRALIZED COMPUTER SERVICES FUNCTION

DISTRIBUTED ORGANIZATIONALSTRUCTURE

27

Characteristics of DDP1

Advantages: Cost reductions in hardware and data entry tasks Improved cost control responsibility Improved user satisfaction because control is closer to user level Backup of data can be improved through use of multiple data

storage sites

Disadvantages Loss of control (organization-wide) Mismanagement of organization-wide resources Hardware/software incompatibility Redundant tasks/data Incompatible tasks may be consolidated Lack of standards

1 DDP = Distributed Data Processing

28

Data are retained in central location. Remote IPUs (workstations) send requests

for data. Central site services needs of remote

workstations. Actual data processing is performed at remote

workstation

Centralized Databases in DDP Environment

29

Data Currency

Can be an issue in DDP with centralized database During transaction processing, data will

temporarily be inconsistent as record is being read and updated.

Database lockout procedures are necessary to keep workstations from reading inconsistent data and/or from writing over a transaction being written by another workstation.

Distributed Databases: Partitioning

Splits central database into segments that are distributed to their primary users

Advantages: users’ control is increased by having data stored at

local sites transaction processing response time is improved volume of transmitted data between workstations is

reduced Reduces potential data loss from a disaster

Distributed Databases: Replication

Duplication of entire database for multiple workstations

Effective for situations with a high degree of data sharing, but no primary user supports read-only queries.

Data traffic between sites is reduced considerably.

32

The Deadlock Phenomenon

Especially a problem with partitioned databases Occurs when multiple sites lock each other out

of data that they are currently using Special software needed to analyze and resolve

conflicts.

The Deadlock Phenomenon

A,BE, F

C,D

Locked A, waiting for C

Locked C, waiting for E

Locked E, waiting for A

34