Anatomy of Business Intelligence

gruesomebugscuffleΛογισμικό & κατασκευή λογ/κού

25 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

332 εμφανίσεις

ISQS 6339, Business Intelligence

Anatomy of Business Intelligence

Zhangxi Lin

Texas Tech University

1

Learning Objectives


Understand the general ideas in business
intelligence by cases


Catch main BI concepts


Get familiar with the most popular BI
applications and tools


Know how to access your BI resources for
this course

2

3

Outline


BI case studies


BI framework


BI applications


BI tools




BI Case Studies


Cases


Premier Bank Card
3’46”


SAP Rail Analytics

9’44”



Advanced applications


Mobile Business Intelligence with iPhone
2’58”


Business intelligence with iPad
6’49”


4

5

Case: Toyota Motor Sales USA


2011
Gartner BI Excellence Awards


Challenge: how to reduce vehicle transit cost


Average: $8/day
-
car * 10 days = $72
-
80/car, 9
-
10days/transit


Total $144
-
160 million/year for transit of two million cars


Problem:


Inability to deliver cars to dealers timely


Computers generated tons of directionless reports and data with little help


Unable to make timely decisions

Case: Toyota Motor Sales USA (II)


Solution: Data warehouse


Use right technologies provided by a right vendor following correct concepts


Oracle’s data warehouse + Hyperion’s BI platform


Lesson learned: data cleansing is important


Results


Discovered that the company was billed twice in some occasions


Increase the volume of cars by 40% between 2001
-
2005


In
-
transit time was reduced 5%


Market share increased


According to IDC Inc. the return on the BI investment was 506%



6

7

Questions


In what way did the old system create problems for
Toyota?


What information needs of managers
re

satisfied by the
new system?


What decisions are supported by the BI system?

8

Case: Chery Automobile


Chery Automobile (simplified
Chinese:
奇瑞汽

) is an
automobile manufacturer in
China.


It is owned by the local
government of Wuhu, Anhui
province, but is scheduled to
be privatized.


In 2009 Chery produced
508,500 units of automobile.


Of the 500,000 Chery vehicles
sold in 2009, 409,300 units
were sedans.


It is the largest independent
Chinese auto manufacturer
and one of the fastest growing
automakers in the world.

Source: http://en.wikipedia.org/wiki/Chery_Automobile

Chery Automobile’s BI System


Problems


supply chain did not
meet JIT


Suppliers have to build their
inventory storages in surrounding
Chery main production lines.


Objectives


Improve the performance of JIT
system


Save the cost of part suppliers by
reducing the inventory storages.


Solution


Hefei University of Technology
helped the company establishing a
BI system to improve the efficiency
of production.


A 90
-
day automatic early
notification of order is set to
coordinate the suppliers’
production plan.


9

Questions


What is the relationship between BI and MIS?


What are main applications in BI?


What are main BI technologies?


10

11

Different Users of Business Intelligence

11


There are many different users who can benefit from
business intelligence


Executives


Those who focus on the overall
business


Business Decision Makers


Usually focused on
single areas of the business (finance, HR,
manufacturing, and so forth)


Information Workers


Typically managers or staff
working in the back office


Line Workers


Employees who might use BI without
knowing it


Analysts


Employees who will perform extensive
data analysis

12

Business Intelligence

BI Users

Executives



Managers




Operators

Data

Business

Environment

Decisions

Data

Business

Analytics

BI Applications:

-

Data warehousing

-
Data mining

-

BPM

-

OLAP

-

etc.

BI vs. MIS


MIS were used by a select few in the organization, due
to the efforts involved in collecting data and preparation
of summaries from the same. Use of Computers in the
MIS process helped in speeding up the process and
increased its reach.


Then came the phase of BI


BI has been privilege of the TOP segment of the pyramid.


The cost involved prohibited the expansion of BI to the Middle tier
Business users.


Operational BI is expected to do the task of making it available to
masses, as they are the ones who need BI, more times in a day.


13

BI Framework

15

A Framework for

Business Intelligence (BI)


The Origins and Drivers of Business
Intelligence


Organizations are being compelled to capture,
understand, and harness their data to support
decision making in order to improve business
operations


Managers need the
right information
at the
right
time
to make
right decisions
.

16

Changing Business Environments

and Computerized Decision Support

17

A Framework for

Business Intelligence (BI)


BI’s Architecture and Components


Data Warehouse


Business Analytics


Automated decision systems


Performance and Strategy

18

A Framework for

Business Intelligence (BI)

19

A Framework for

Business Intelligence

20

A Framework for

Business Intelligence (BI)


BI’s Architecture and Components


Data Mining


A class of information analysis based on
databases that looks for hidden patterns in a
collection of data which can be used to predict
future behavior

21

A Framework for

Business Intelligence (BI)


BI’s Architecture and Components


business (or corporate) performance
management (BPM)



A component of BI based on the
balanced
scorecard
methodology, which is a framework for
defining, implementing, and managing an
enterprise’s business strategy by linking
objectives with factual measures

22

A Framework for

Business Intelligence (BI)


BI’s Architecture and Components


User Interface: Dashboards and Other
Information Broadcasting Tools


Dashboards


A visual presentation of critical data for executives to
view. It allows executives to see hot spots in seconds
and explore the situation


23

A Framework for

Business Intelligence (BI)


The Business Value of BI


How BI Can Help


Assess their readiness for meeting the challenges posed by
these new business realities


Take a holistic approach to BI functionality


Leverage best practices and anticipate hidden costs


Key Issues and Framework for BI Analysis


How can enterprises maximize their BI investments?


What BI functionality do enterprises need, and what are they
using today?


What are some of the hidden costs associated with BI
initiatives?


24

A Framework for

Business Intelligence (BI)


Time savings


Single version of truth


Improved strategies and
plans


Improved tactical
decisions


More efficient processes


Cost savings


Faster, more accurate
reporting


Improved decision making


Improved customer service


Increased revenue


The Benefits of BI

Hadoop


for BI in the Cloud era


Hadoop is a free, Java
-
based programming framework that supports
the processing of large data sets in a distributed computing
environment.


Hadoop makes it possible to run applications on systems with
thousands of nodes involving thousands of

terabyte
s.


Hadoop was inspired by

Google
's

MapReduce
, a software
framework in which an
application

is broken down into numerous
small parts. Doug Cutting, Hadoop's creator, named the framework
after his child's stuffed toy elephant.


The current Apache Hadoop ecosystem consists of the
Hadoop

kernel,

MapReduce, the Hadoop

distributed file
system
(HDFS) and a number of related projects such as

Apache
Hive
, HBase and

Zookeeper
.


25

Hadoop & MapReduce

26

MapReduce is a framework for processing

parallelizable


problems across huge datasets using a large number of

computers (nodes), collectively referred to as a cluster


or a

grid.


Hadoop
Architecture


Hadoop consists of
the

Hadoop
Common

which provides
access to the filesystems
supported by Hadoop.
The Hadoop Common
package contains the
necessary JAR

files and
scripts needed to start
Hadoop. The package
also provides source
code, documentation,
and a contribution
section which includes
projects from the
Hadoop Community.

27

Dell representation of the Hadoop
ecosystem

28

BI Applications

30

ISQS 3358 Business Intelligence

30

Main BI Topics

30


Data warehousing


Making historical data available for analytics


Data preparation


Extraction, transformation and loading


Query
-

a collection of specifications that enables you to focus on
a particular set of data.


Online Analytical Processing (
OLAP
)
-

a capability of information
systems that supports interactive examination of large amounts
of data from many perspectives.


Reporting
-

generates aggregated views of data to keep the
management informed about the state of their business.


Data mining
-

extraction of knowledge by utilizing software that
can isolate and identify previously unknown patterns or trends in
large amounts of data.

31

BI Applications


Scorecards


Dashboard


Reports


Analytic applications


Data mining


Predictions


OLAP & OLTP


32

Business Scorecards

32

We see this
everyday

33

The Purpose of a Scorecard

33


A scorecard should give an executive a visual
representation of the health of an organization
in a single glance


The scorecard is of sufficiently high level to
represent major business operations and their
goals


The data in a scorecard should be as recent as
possible to make them more actionable

34

Benefits to the Executive

34


In a single glance, the executive can see a wide
swath of the business (finance, manufacturing,
sales, marketing, and more)


Immediate value is gained without the need for
the executive to perform analysis


Executives see not just actual values, but
comparisons to plans or prior results

35

The Contents of a Scorecard

35


Scorecards usually contain some or all of the
following elements:


Key Performance Indicators (KPIs)


KPI actual values compared to historical values (for trend
analysis)


KPI actual values compared to a forecast or budget
amount


Rankings of different departments, locations, products, and
so forth



Developing KPIs and Scorecards with SharePoint

36

Dashboards

36

37

The Purpose of a Dashboard

37


A dashboard is designed to allow decision
makers to see a variety of data that affects their
divisions or departments


This data may be in the form of scorecards,
charts, tables, and so forth


The dashboard is generally customized for
each user


More targeted and detailed than a scorecard

38

Benefits to Decision Makers

38


Decision makers see a variety of information
targeted to their department


This allows decision makers to focus only on
the items over which they have control


Information is more detailed than that of a
scorecard


The tools in the dashboard often have better
analytic capabilities than a scorecard

39

The Contents of a Dashboard

39


A Dashboard generally contains a variety of
different views of data


The data is generally KPIs and shows trends,
breakdowns, and comparisons against a
forecast or historical data


The dashboard often consists of charts and
tables, and may include scorecard elements as
well

40

Reports

40

41

The Purpose of Reports

41


Reports allow a much broader audience to benefit
from the data in a BI solution


Reports may be static, requiring no training


Reports may also allow a limited amount of
interactivity


Reports can be presented in a variety of formats,
allowing for easier distribution

42

The Purpose of Custom Application
Integration

42


An application used by line workers may include
business intelligence without the worker
realizing what is happening


A sales clerk may get a list of targeted
recommendations to make based on what the
customer is buying


A loan officer may be presented with the level
of risk associated with granting a loan to a
particular customer

43

The Contents of Custom Application
Integration

43


Custom applications may include predictive
output from data mining models


Custom applications can show history and
trends for the current customer, supplier, and so
forth


Custom applications may allow easy ways for
users to explore the data for relationships

44

Analytic Applications

44

45

The Purpose of Analytic Applications

45


Analytic applications free analysts from building
complex models and writing complex queries


Analysts are free to focus on the data and
discover relationships and drivers behind
numbers


Rich visualizations allow much easier
understanding of trends and relationships

46

The Contents of Analytic Applications

46


Analytic applications typically have no limits;
analysts can see everything


Analytic applications can view and analyze all of
an organization’s data in a number of ways


Analytic applications are powerful, but not as
easy to use as other mechanisms

47

ISQS 3358 Business Intelligence

47

OLTP vs. OLAP


Online transaction processing systems (OLTP)


Systems that

handle a company’s routine ongoing
business


Online analytic processing
(OLAP)


An information system that enables the user, while at a
PC, to query the system, conduct an analysis, and so on.
The result is generated in seconds

48

Case: A Clustering Example

ID

Name

Gender

Age

Income

C001

X

M

15

2000

C002

Y

F

31

5500

C003

Z

M

18

3400

C004

A

F

45

3500

C005

B

M

30

3000

C006

C

F

25

4000


The data is used to
segment the
customers for sale
promotion


Three products:
DVD, game, drink for
adult


Problems


How to segment
the customers into
several clusters


Are these clusters
good enough?



Data & Text Mining

48

49

Case:
AOL search data scandal (2006)


AOL


On August 4, 2006, AOL Research released a compressed text
file on one of its websites containing twenty million search
keywords for over 650,000 users over a 3
-
month period, intended
for research purposes, which was posted only three days before
pulled down.


While none of the records on the file are personally identifiable,
the New York Times was able to locate an individual from the
released and anonymized search records by cross referencing
them with phonebooks or other public records.


Questions:


Is it ethically fine to collect data about the rivals?


How to protect the confident data from the espionage of
competing counterparts?


50

AOL Search Data Case 1


User A:


Searched for a home in the Pine Tree state.


Between the searches, also looked for


"how to make home made bombs,"


"how to make a meth lab," and


"www.morbiddeaths.com."


Also interested in Stephen King


Possibility: User A is to move to Pine Tree state to
assassinate Stephen King with a home made bomb

51

AOL Search Data Case 2


User B


Has a taste for kiddie porn, because of searches for "lolita nudes,"
"underground kiddy porn pictures," "russian preteen nudes," and,
"see girls and animals having sex."


Appears to own a 2004 Corvette because "need oil drain plug gasket
for my 2004 corvette"


Lives near Baltimore because "see telephone number of old wheel
world in Baltimore".


Planning a trip to Vegas because "see the weather in las vegas" and
"see showtime tours in las vegas".


This set of information could be enough to narrow down who he is.



Some extra information about the person can narrow down who
he is.

2011 China Social Media Password Leak


In 2011, a huge cache of personal data from China’s most popular
websites leaked onto Internet. Between December 21 and 25,
hackers released more than 100 million users account information
including usernames, passwords, and emails (reference here).


Victims include users in IT technical community, social networking,
gaming, and microblogging websites.


Due to many China’s online companies failed to encrypt user
password, the online security crisis has caused panic among
Internet users in China and worry about more websites were hacked
and personal information is at risk.


52

Password Length


Except for company B, the
average password lengths in
companies A, C, and D are
close to 8 digits.


The average password length
of company B is 9.46 (about
1.5 digits higher than users in
A, C, D).


The results indicate IT
professions have higher sense
of security even the company
B has lower password length
requirement.

A

B

C

D

Average

Password

Length

7.94

9.46

7.96

7.80

53

Unsecured Rules of Password
Constitution


The results show most
long password
constitutions followed
one of the rules below:


Name abbreviation + ID
numbers


Full name + phone
numbers


English name + birthday


Full email address


Pingyin (transcribe
Chinese characters into
the Roman alphabets)


54

BI Tools

56

ISQS 3358 Business Intelligence

56

BI Product Providers

56


Microsoft


SAS


IBM


Oracle


SyBase


Business Objects



BI Tools Survey


57

ISQS 3358 Business Intelligence

57

List of BI tools

57

No.

Tool

Version

Vendor

1.

Oracle Enterprise BI Server

7.8

Oracle

2.

Business Objects Enterprise

XI r2

Business Objects (now
SAP)

3.

SAP NetWeaver BI

7.0

SAP

4.

SAS Enterprise BI Server

9.1.3

SAS Institute

5.

TM/1 & Executive Viewer

9.1

Applix (now IBM)

6.

BizzScore Suite

7.2

EFM Software

7.

WebFocus

7

Information Builders

8.

Excel, Performance Point,
Analysis Server

2007/2005

Microsoft

9.

QlikView

8

QlikTech

10.

Microstrategy

8

Microstrategy

11.

Hyperion System

9

Hyperion (now Oracle)

12.

Actuate

9.1

Actuate

13.

Cognos Series 8

8.3

Cognos (now IBM)

58

Software Used in this Class

ISQS 6339, Data Mgmt & BI

58



Microsoft SQL Server 2008


SAS Enterprise Guide v4.2


Base SAS for Data Preparation
Programming



59

Conclusion:

Business Intelligence Today and Tomorrow


Today’s organizations are deriving more value from BI
by extending actionable information to many types of
employees, maximizing the use of existing data assets


Visualization tools including dashboards are used by
producers, retailers, governments, and special agencies


More and more industry
-
specific analytical tools will flood
the market to perform almost any kind of analysis and to
facilitate informed decision making from the top level to
the user level


A potential trend involving BI is its possible
merger with
artificial intelligence
(AI)


Reading Assignments


Find and read “
How Much Information? 2003

(
http://www2.sims.berkeley.edu/research/projects/how
-
much
-
info
-
2003/execsum.htm
)
-

Could you find newer
information about the information explosion?


Find a BI application case from the web, and understand
how it works.


Find paper “CACM2011 Overview of BI.pdf” in the
network drive under ~
\
Texts
\
Readings
\
. Read it carefully.

60

61

CAABI

ISQS 6339, Data Mgmt & BI

61


Center for Advanced Analytics and Business
Intelligence

initially started in 2004 by Dr. Peter
Westfall, ISQS, Rawls College of Business.


Looking to offer support to companies in developing
BI capabilities.


Lots of technical expertise.