Transcript Lecture 16.pptx

Managing Information
Resources
Lecture 16
Managing Information Resources

Managing Information Resources Lectures explores the
management of data information, and knowledge

It begins by identifying some problems in managing data,
and then surveys the evolution of database management
systems, including the next-generation systems

It explores the various types of information that
companies need to manage as they treat information as
an organizational resource
Managing Information
Resources

It concludes by discussing one of the most important
issues facing companies today: how to manage
knowledge

Case examples include Monsanto, Owens & Minor,
HICSS Personal Proceedings, Tapiola Insurance Group,
Tennessee Valley Authority, Eastman Chemical
Company and Groove Networks
Today’s Lecture


Introduction
Managing Data
 The
Three-Level Database Model
 Four
Data Models
 Getting
Corporate Data into Shape
Today’s Lecture..

Managing Information
 Four
Types of Information
 Data
Warehouses
 Document
 Content
Management
Management
Introduction

“Managing information resources” initially meant
managing data, first in files, then in corporate databases
which were:
 Well
structured
 Carefully
defined, and
 Controlled
by IS department
Introduction

Data vs. Information vs. Knowledge
 Data:

facts devoid of meaning or intent
 Information:
data in context
 Knowledge:
information with direction or intent
As the breadth of the kinds of information resources has
expanded, so has the job of managing them. The job
may not start in the IS department but it invariably ends
up there
Introduction
 PCs
users used ‘alone’

Needed to share files

Version control, back-up etc.
 Web
sites / content
Introduction

Initially created their own

Need for recovery, version control

Corporate consistency
 IS
to the ‘rescue’

Management procedures

Discipline
Introduction ..

Corporate databases are still a major IS department
responsibility
 Sometimes
housed in a variety of database models
 Production
databases – transaction
 Data
warehouses
 CRM
–Customer Relationship Management
Introduction

Information in the form of documents (electronic or
paper) and Web content has exploded the size of
databases organizations now manage

Knowledge management is becoming a key to exploiting
“intellectual assets”

Information resources need to be well managed as
information becomes an important strategic resource
Managing Data

Database management systems are the main tool for
managing computerized corporate data

They have been around since the 1960s and are based
on two major principles:
 A three- level conceptual model and
 Several alternative ‘data models’ for organizing the
data
DBMS EXAMPLE
Managing Data:
The Three-Level Database Model

Level 1 – The external, conceptual, or local level,
containing the various “user views” of the corporate data
that each application program uses
 Not
concerned with how the data will be physically
stored or what data is used by other applications

Managing Data:
The Three-Level Database Model

Level 2 - The logical or “enterprise data” level
 ‘Technical’
(human) view of the database = under
control of the DBAs

Level 3 - The physical or storage level, specifying the
way the data is physically stored
 End
user not concerned with all these ‘pointers and
flags’ (how the data is physically organized) = they
are for use by the DBMS
The Three-Level Database Model:
Advantages

Level 2 absorbs changes made at Level 3 such
as using a new physical storage device
 Individual
application programs in Level 1 do not need
to be changed when the physical layer changes

Data only needs to be stored once in Level 2,
and different programs can draw on it and vary
the relationships among the data
Managing Data:
Four Data Models
The second major concept in database management is
alternate ways to define relationships among data
Hierarchical model: structures data so that each
element is subordinate to another in a strict
hierarchical manner
1.
‒
Parent, child etc.
Network model: allows each data item to have more
than one parent,
2.
‒
Relationships stated by pointers stored with the data
Managing Data:
Four Data Models cont.
3.
Relational model:
where the data is stored in tables.
–
Eight relational operations can be performed on this
data
Select, Project, Join, Product, Intersection,
Difference, Union, Division
4.
Entity-Relationship model:
Managing Data:
Four Data Models cont.


Microsoft Access
Relational systems are not as efficient as hierarchical or
network database systems, but because relational systems
allow people to create relationships among data on the fly.
More flexible
Microsoft Access DBMS Example
Managing Data:
Four Data Models cont.


First used to handle end user queries – they are now
widely used in high-volume transaction systems with
huge files
Hence, they have become the database technology of
choice in today’s systems

Also = largely due to decrease in costs of
technology: processing and disk storage
Managing Data:
Four Data Models cont.
Object model: can be used to store any type of data, whether a:
4.
–
Traditional name or address,
–
An entire spreadsheet,
–
A video clip,
–
A voice annotation,
–
A photograph, or
–
A segment of music
Managing Data:
Four Data Models cont.
The tenets of objects have become increasingly
important in the world of computing

–
E.g. Web Services because the XML modules
utilize object principles
Managing Data:
Four Data Models cont.
Typical, yet complex database applications that may
require objects:

–
CAD for a large office building
–
Large retail chains record every product code
scanned
–
Insurance policy files e.g. claim forms, images,
video etc.
Managing Data:
Four Data Models cont.

Object models retain traditional DBMS features
including:





End user tools
High level Query languages
Concurrency control
Recovery
Ability to handle huge amounts of data
Managing Data:
Four Data Models cont.

Include two other major concepts
 Object

management
Management of complex kinds of data such as
multimedia and procedures
 Knowledge

management
Management of large numbers of complex rules for
reasoning and maintaining integrity constraints
between data
Managing Data:
Four Data Models cont.

Finally, security is of major importance in today’s DBMSs
 Problem
= compounded by distributed,
heterogeneous Internet-linked databases

Companies may want to permit access to some portions
of their databases whilst restricting other portions
Managing Data:
Four Data Models cont.
 This
selective accessibility requires reliably
authenticating ‘users’

Unless security and integrity are strictly enforced, users
will not be able to (fully) trust the systems
Managing Data

Getting Corporate Data into Shape
Getting Corporate Data into Shape

In the midst of this growing richness of data and
information, companies are still struggling to get their
internal alphanumeric data under control

The installation of company-wide software packages
such as SAP, enterprise data warehouses, and intranets
has once again brought to the fore the problems of “dirty
data”
Getting Corporate Data into Shape
 Data
from different databases that has:
 Different names
 Uses different time frames, or
 That otherwise does not match

Attempts to get under control go back a long way:
 Late
’60s / early ’70s = DBMS
 Then
= the still evolving and important role of “data
administration:
 Managing all the computerized data resources of a
company
Getting Corporate Data into Shape: The
Problem: Inconsistent Data Definitions

Problem: data definitions incompatible from:
 Application
to application
 Department
 Site
to department
to site, and
 Division
to division
Getting Corporate Data into Shape: The
Problem: Inconsistent Data Definitions

Reason: to get application systems up and running
quickly, system designers sought data from the cheapest
source or politically expedient source

Result: different files with:
 Different
 Same
names for same data, and
name for different data etc
Getting Corporate Data into Shape: The
Problem: Inconsistent Data Definitions cont.






Account Number
AcctNum
AcctNumb
Acct#
A/CNum
Note: people (in the majority of cases) weren’t stupid
 They never dreamt their files / databases etc. would
be used in this manner
 Historical ‘stand alone’ computing
 Information collation, use, communication etc. =
never thought possible
Getting Corporate Data into Shape: The
Role of Data Administration

The use of DBMS - database management software,
reduced, to some extent, the problems of inconsistent
and redundant data in organizations
 However
merely installing & running a DBMS is not
sufficient to manage data as a corporate resource

Database administration: concentrates on administering
databases and the software that manages them
Getting Corporate Data into Shape: The
Role of Data Administration cont.

Data administration is broader:
 To
determine what data is being used outside the
organizational unit that creates it
 Whenever data crosses organizational boundaries, its
definition and format need to be standardized

Data dictionaries are the main tools by which
data administrators control standard data
definitions
Getting Corporate Data into Shape:
ERP (Enterprise Resource Planning)

To bring order to the data mess, data administration
has four main functions:
1. Clean up the data definitions
2.
Control shared data
3.
Manage data distribution, and
4.
Maintain data quality
Getting Corporate Data into Shape:
ERP (Enterprise Resource Planning)

Interestingly, many companies really did not take these
four jobs seriously until the mid 1990s, when they
needed consistent data to install a company-wide ERP
package

ERP provided the means to consolidate data to give
management a corporate-wide view of operations
Monsanto
Case Example: Managing Corporate Data / ERP


Monsanto case study to illustrate one company’s
success in getting its corporate data in shape
Monsanto is a provider of agricultural products,
pharmaceuticals, food ingredients, and chemicals.
50% revenues outside USA, it is decentralized
Monsanto
Case Example: Managing Corporate Data / ERP

Monsanto established three large enterprise wide IT
projects:
1.
To redevelop operational and financial transaction
systems using SAP
2.
To develop a knowledge-management architecture,
including data warehousing, and
3.
To link transaction and decision support systems via
common master data, known as enterprise
reference data (ERD)
Monsanto
Case Example: Managing Corporate Data / ERP
cont.

Monsanto is too large and complex to operate SAP as a
single installation
 They
 With
have created a distributed SAP architecture
separate instances of SAP for reference data,
finance, and operations in each business unit
 The master reference data integrates these
distributed components
Monsanto
Case Example: Managing Corporate Data / ERP
cont.

To convert SAP data to knowledge, Monsanto uses data
warehouses
 The
sole source of master data is the ERD, but the
data can be distributed wherever they are needed

To get corporate data in shape, Monsanto created a
department called ERD Stewardship to set data
standards and enforce quality—hence its nickname, “the
data police.”
 Independent
of MIS
Monsanto
Case Example: Managing Corporate Data / ERP
cont.

Another newly created function is entity specialists =
managers with the greatest stake in the quality of data

Also, data managers who now adhere to the new ERD
rules
 This has led to a cultural change: The idea of
“tweaking” a system to fix a local discrepancy,
formerly common, can now cause a major disruption
in operations or a bad decision based on faulty data
Monsanto
Case Example: Managing Corporate Data / ERP
cont.

Getting the data in shape was a huge undertaking, but it
has made the company more flexible

Monsanto is already reaping bottom-line benefits from
better integration and greater flexibility
Summary

Introduction

Managing Data
 The Three-Level Database Model
 The
Three-Level Database Model:
 Four
Data Models
Advantages
Summary..


We have covered today
Getting Corporate Data into Shape

The Problem: Inconsistent Data Definitions
 The

Role of Data Administration.
 ERP (Enterprise Resource Planning)
Monsanto
Case Example: Managing Corporate Data / ERP