Decision Support Systems

1

CHAPTER 10

DATA, KNOWLEDGE,

AND DECISION SUPPORT

2

Management and

Decision Making


Management


a process by which certain goals are achieved
through the use of resources


Two main phases for decision making:


Problem identification and possible solutions
formulation: information filtration, analysis,
and interpretation


Choice of appropriate solution

3

Reasons for IT support


The increasing number of alternatives


Time pressure


Decision complexity


The need to access remote information and
knowledge


4

The Data Life
-
cycle Process


Business needs information and knowledge


Data collection


Data transformation:


Data storage and management


Data analysis and processing


Document management


Knowledge management

5

The Data Life Cycle Process

Data Sources
(databases)

End Users:

Decision Making and other Tasks;

Data Visualization

Data Warehouse

(storage)

Analytical Processing,

Data Mining

Generate
Knowledge

Organizational
Knowledge Bases

Purchased
Knowledge

Storage

Direct
Use

Data
Organization;
Storage

Direct Use

Use

Use

Use

Use of
Knowledge

Storage

6


Internal Data

-

generated within organization by
the corporate TPS, FIS, MIS


Personal Data
-

created by IS users or other
corporate employees documenting their own expertise


External Data
-

generated outside and organization


Methods for Collecting Raw Data


manually or by instruments and sensors


transferred electronically

Data Sources and Collection

7

•
Accurate

•
Secure

•
Relevant

•
Timely

•
Complete

•
Consistent


Data Quality


8


Exponential increases of data with time


Various sources of raw data


Only small portions are relevant


Increasing amount of external data


Different legal requirements relating to data


Selecting data management tools
-

a problem


Data security, quality, and integrity

Difficulties in data management

9

Decision Making Process



Intelligence Phase

Design Phase

Choice Phases



REALITY

Implementation

of Solution



SUCCESS

FAILURE

Testing of Proposed
Solution

Verification
of the Model

Examination


10


A model is a simplified representation of reality


Models classification:


Mental

model (or conceptual model) is verbal
description of reality


Iconic
(scale) model is a physical replica of a
system, usually based on a different scale form
original


Analog

model
-

a physical model, but the shape of
the model differs from that of the actual system


Mathematical

(quantitative) model describes a real
system based on mathematical formulas and
constructions

Models

11

Mathematical Models


Model variables
–

investigated
characteristics of real world system


Parameters
–

represent internal and external
conditions


Managerial solutions are reflected in
model’s initial values and parameters


12

Mathematical models (cont….)


Analytical models


Simulation models:


Advanced math. techniques


Computational methods


Computational algorithms


IT support

13

Model Investigation


Model validation


Stability analysis
–

model reaction to small
disturbances in initial values


Sensitivity analysis
–

model reaction to
small disturbances on parameters values


Simulation experiments

14

Simulation experiments


What
-
if analysis
–

checks the consequences
of possible solution


Goal
-
seeking analysis
–

attempts to find
inverse solution:


Not every model has inverse solutions


Computational algorithms based on series of
direct simulations must be used

15


Structured decision making process
–

all four
stages are structured


Semistructured decision making process
–

not
all stages are structured


Unstructured decision making process
–

all four
stages are unstructured, required intuition and
knowledge


A Framework for Computerized
Decision Support

16


Systematic process for solving problems


Define the problem


Classify the problem into a standard category


Construct a standard mathematical model


Find potential solutions


Choose and recommend a specific solution

Management Science

17



support decision makers at all managerial levels


support several interdependent and/or sequential
decisions


support all phases of decision making and variety of
decision
-
making processes and styles


can be adapted over time to deal with changing
conditions


utilize models


integrate systems


execute analysis of models

DSS Characteristics

18

Are implemented on the software level


Data Management


User Interface


Model Management


Knowledge Management


Users


Components of DSS

19

Decision Support Systems


Individual DSSs:


Functional analysts


Low
-
level managers


Group Decision Support Systems (GDSS):


Groups of managers


Top
-
level managers

20

Group Decision Support
Systems


Specially designed


User
-
friendly


Flexible


Support collaboration of geographically
dispersed users


Contain nominal group techniques:


Send feedback




Votes


Anonymous inputs




Keeping records

21

Executive Information Support


Drill down


Critical success factors and key performance indicators


Status access


Access to the external information and knowledge


Trend analysis


Ad hoc analysis


Exception reporting


Integration with DSS

22

Data Visualization Technologies

Present data in clear and understandable
form


Traditional forms:


digital images, graphs, charts, animation,
multimedia


Visual Interactive Decision Making


Visual interactive modeling


Geographical Information Systems

23

Geographical Information
systems


DSSs supporting decision making process
using digital maps


Contain geographically referenced data
tying to objects on a map


Databases, spreadsheets, analytical tools
and user interface are main components

24

Geographical Information
System (GIS)

GIS

Spatial Imaging
Function

Design and
Planning
Function

Database
Management
Function

Decision Modelling
Function

Surveying and Mapping

Design and
Engineering

Facilities
Management

Strategic
Planning and
Decision Making

Demographic
and Market
Analysis

Transportation and Logistics

25

Emerging GIS Applications


help reengineer the aviation,
transportation, and shipping industries


enables vehicles or aircraft equipped with
a GPS receiver to pinpoint their location
as they move


include railroad car tracking and earth
-
moving equipment tracking

26

Knowledge Management

Knowledge capturing, storing,
distribution require:


Knowledge identification


Knowledge discovery and analysis


Establishing organizational
Knowledge base



27

Types of Organizational
Knowledge


Knowledge assets
-

regarding markets, products,
technologies, and organizations that a business owns
or needs to own


Best practices
-

collection of the most successful
solutions and/or case studies


Intellectual capital
-

collection of knowledge
amassed by an organization over the years


competitive intelligence
-

collection of
competitive information


28


Identify valid, novel, potentially useful
data, and understand patterns in data


Supported by : massive data collection,
powerful multiprocessor computers, and
data mining and OLAP algorithms


Tools : data mining and online analytical
processing

Knowledge Discovery

29


Analysis by end users from their desktop, online


Analyze the relationships between many types of
business elements


Involve aggregated and summarized data


Compare data over hierarchical time period


Present data in different perspectives


Work with queries

Online Analytical Processing
(OLAP)

30


Data Mining searches for valuable
business information in a large database
and “mines a mountain for a vein of
valuable ore”


Functions:


Classification


Forecasting


Clustering



Association


Sequencing




Data Mining for Decision
Support

31

Data Mining
vs.
OLAP

OLAP

Data Mining

Purpose

Supports data
analysis and
decision making

Supports data
analysis and
decision making

Types of analysis

Top
-
down,
query
-
driven

Bottom
-
up,
discovery
-
driven

User’s Skills

Data analysis
and data
business context

Must trust data
mining tools