Transcript Data Mining Life Cycle

Business Intelligence
and Data Mining
Session 14 Section 4
Instructor: Michael Sutton, PhD, CMC, AdmA, MIT
EMBA 512
Fall 2015
 Enterprise Performance
Management (EPM)
 BVP of BI Presentation Layers
for EPM (Key Characteristics
and Examples)
Agenda




Dashboards
Visual Analysis Tools
Scorecards
Reports
 Data Mining & Business
Intelligence
 Conceptual Model & Life
Cycle




Classification
Prediction
Clustering
Mining Association Rules
Boise State Executive MBA Program Fall 2015
 Data Mining Life Cycle
 SEMMA
 CRISP
 How Can Business
Intelligence Drive Profits?
 Best Practices: BI Project
Management Life Cycle:
 Best Practices: Integrate
KPIs
 Future Opportunities for BI,
DW and DMg
2
http://pelotongroup.com/what-we-do/enterprise-performance-management/
Enterprise
Performance
Management
(EPM) [1]
Boise State Executive MBA Program Fall 2015
3
Enterprise
Performance
Management
(EPM) [2]
 EPM builds upon BI by leveraging information after implementing
a BI framework—organizations move to automate processes that
leverage this information.
 An EPM system integrates and analyzes data from many sources,
including, but not limited to, e-commerce systems, front-office
and back-office applications, data warehouses and external data
sources.
 Advanced EPM systems can support many performance
methodologies such as:




dashboards
visual analytical tools
scorecards
reports.
Boise State Executive MBA Program Fall 2015
http://www.gartner.com/it-glossary/epm-enterprise-performance-management
 Enterprise Performance Management (EPM) is the process of
monitoring performance across the enterprise with the goal of
improving business performance.
4
BVP of BI
Presentation
Layers for EPM
 nothing but performance indicators behind GUIs (Graphical User
Interface).
 effectiveness is due to a careful selection of the relevant measures,
while highlighting data quality standards.
 not the primary goal of a BI/DW System.
 viewed as a sophisticated and effective add-ons to a BI/DW System.
 Primary goal of a BI/DW System should always be to properly
define a process to transform data into actionable information
(knowledge “between the ears” of the executives and managers).
Boise State Executive MBA Program Fall 2015
Golfarelli, Rizzi, and Cella (2004), Beyond Data Warehousing: What’s Next in Business Intelligence?
 Presentation Layers:
5
https://www.geckoboard.com/blog/building-great-dashboards-6golden-rules-to-successful-dashboard-design/#.VjmIlrerRAk
NOT a
BI/BA
Dashboard!
Boise State Executive MBA Program Fall 2015
6
All the visualizations fit on a single computer screen — scrolling to see
more violates the definition of a dashboard.
2.
Displays the most important (key) performance indicators / performance
measures to be monitored.
3.
Interactivity such as filtering and drill-down can be used; however, those
types of actions should not be required to see which performance
indicators are under performing.
4.
Not designed exclusively for executives, but rather should be used by the
general workforce staff and managers, thus easy to understand and use.
5.
Displayed data automatically updates without any assistance from the
user. The frequency of the update will vary by organization and by purpose.
The most effective dashboards have data updated at least on a daily basis.
6.
Operational dashboards: (http://www.klipfolio.com/resources/dashboard-examples)
 manage intra-daily business processes – frequently changing and current
performance metrics or key performance indicators
7.
Analytical dashboards (http://www.klipfolio.com/resources/dashboard-examples)
 focus on gaining insights from a volume of data collected over time – often the
past month or quarter – and use this to understand what happened, why, and
what changes should be made in the future.
Boise State Executive MBA Program Fall 2015
http://www.dashboardinsight.com/articles/digitaldashboards/fundamentals/what-is-a-dashboard.aspx#sthash.wry24hel.dpuf
Key
Characteristics
of EPM
Dashboards
1.
7
http://www.dashboardinsight.com/CMS/d7e568f7-ac5b-4a809150-54c017548473/business-dashboard-example.png
Example of
EPM
Dashboards
Boise State Executive MBA Program Fall 2015
8
Offer the ability to select various date ranges, pick different
products, or drill down to more detailed data.
2.
Fits on one screen, but there may be scroll bars for tables with
too many rows or charts with too many data points.
3.
Highly interactive and usually provides functionality like
filtering and drill downs.
4.
Primarily used to find correlations, trends, outliers (anomalies),
patterns, and business conditions in data.
5.
Generally historical data. However, there are some cases where
real-time data is analyzed.
6.
Identifies key performance indicators for use in dashboards.
7.
Typically relied on by technically savvy users like data analysts
and researchers.
Boise State Executive MBA Program Fall 2015
See more at: http://www.dashboardinsight.com/articles/digitaldashboards/fundamentals/what-is-a-dashboard.aspx#sthash.wry24hel.dpuf
Key
Characteristics
of EPM Visual
Analysis Tools
1.
9
http://www.dashboardinsight.com/CMS/d7e568f7-ac5b-4a80-915054c017548473/public-technology-performance-dashboard.png
Example of
EPM Visual
Analysis Tools
Boise State Executive MBA Program Fall 2015
10
Scorecards and dashboards are often used interchangeably, but
can be differentiated.
2.
Tabular visualization of measures and their respective targets
with visual indicators to see how each measure is performing
against their targets at a glance.
3.
Contains at least a measure, its value, its target, and a visual
indication of the status (e.g. a circular traffic light that is green
for good, yellow for warning, and red for bad) on each row.
4.
Scorecard should not be interactive nor contain scroll bars.
5.
It may contain columns that show trends in sparklines.
Boise State Executive MBA Program Fall 2015
See more at: http://www.dashboardinsight.com/articles/digitaldashboards/fundamentals/what-is-a-dashboard.aspx#sthash.wry24hel.dpuf
Key
Characteristics
of EPM
Scorecards
1.
11
http://www.dashboardinsight.com/CMS/d7e568f7-ac5b-4a809150-54c017548473/dashboard-design-scorecard-example.png
Example of
EPM
Scorecards
Boise State Executive MBA Program Fall 2015
12
Reports contain detailed data in a tabular format and typically
display numbers and text only, but they can use visualizations
to highlight key data.
2.
Presents numbers and text in a table.
3.
May contain visualizations but only used to highlight findings in
the data.
4.
Optimized for printing and exporting to a digital document
format such as Word or PDF.
5.
Geared towards people who prefer to read data, for example,
lawyers, who would rather read text over interpreting
visualizations, and accountants, who are comfortable working
with raw numbers.
Boise State Executive MBA Program Fall 2015
See more at: http://www.dashboardinsight.com/articles/digitaldashboards/fundamentals/what-is-a-dashboard.aspx#sthash.wry24hel.dpuf
Key
Characteristics
of EPM
Reports
1.
13
http://www.mrexcel.com/forum/powerbi/788960-profit-loss-powerpivot.html
Example of
EPM Reports
Boise State Executive MBA Program Fall 2015
14
Exercise # 4:
20 min.
EPM
Presentation
Layers [1]
 12 Standard Screen Patterns
 http://designingwebinterfaces
.com/designing-webinterfaces-12-screen-patterns
 30 Essential Controls
 http://designingwebinterfaces
.com/essential_controls
 15 Common Component
Patterns
https://www.pinterest.com/pin/69524387974322158/
 Designing Web Interfaces Principles and Patterns for
Rich Interaction by Bill Scott &
Theresa Neil
 http://designingwebinterfaces
.com/15-common-components
Boise State Executive MBA Program Fall 2015
15
 Break into Groups of 4-6
 Move to Breakout Rooms
Exercise # 4:
20 min.
EPM
Presentation
Layers [2]
 Business Problem:
 Each Team is provided with 4 different examples of BI Presentation
Layers
 Using what you have just learned about Dashboards, Visual Analysis
Tools, Scorecards, and Reports
 Critique each presentation layer
 Suggest improvements
 Identify 1 tool that can be used to create each Presentation Layer.
 Report out
 Teams: Select a spokesperson
 Present and defend results.
Boise State Executive MBA Program Fall 2015
16
Data mining is “non-trivial extraction of implicit, previously
unknown, and potentially useful information from data in
databases” [including DWs].
—Frawley, Piatetsky-Shapiro, Matheus (1991), Knowledge Discovery in Databases: An Overview
http://www.sas.com/en_us/insights/analytics/datamining.html?gclid=CIer3O_q9MgCFVJlfgodxa0ORA
Data Mining—
Advanced Business
Analytics
Alternative synonyms:
Knowledge discovery (mining) in databases (KDD), knowledge extraction, data/pattern analysis,
data archeology, data dredging, information harvesting, data forensics.
Boise State Executive MBA Program Fall 2015
17
http://www.slideshare.net/salahecom/introduction-to-data-mining-tutorial
Data Mining
and Business
Intelligence [1]
Boise State Executive MBA Program Fall 2015
18
http://www.saedsayad.com/data_mining_map.htm
Data Mining:
Conceptual
Model & Life
Cycle [Part 1]
Boise State Executive MBA Program Fall 2015
19
http://www.saedsayad.com/data_mining_map.htm
Data Mining:
Conceptual
Model & Life
Cycle [Part 2]
Boise State Executive MBA Program Fall 2015
20
Classification
[1]
 Classifies data by
constructing a model,
based upon a training set
and the values within a
classifying attribute,
using the value to classify
new data





Terrorist profiling
Target marking
Credit Approval
Medical diagnosis
Fraud detection
Boise State Executive MBA Program Fall 2015
Han and Kamber, (2006). Data Mining: Concepts and Techniques
 Predefined set of groups
or models based on
predicted values.
21
http://www.slideshare.net/salahecom/08-classbasic
Classification
[2]
Boise State Executive MBA Program Fall 2015
22
 Relates to time series; but
not time bound.
Prediction
 Used to predict value
based on past data and
current data.
 Water flow of a river
will be calculated by
various monitors at
different levels and
different time intervals.
Information use to
predict the future
water flow.
Boise State Executive MBA Program Fall 2015
https://www.siggraph.org/education/materials/H Han and Kamber, (2006). Data Mining: Concepts and Techniques
yperVis/applicat/data_mining/data_mining.html
 Similar to Classification
23
 similar (or related) to one another within the same group
 dissimilar (or unrelated) to the objects in other groups
 Clustering (cluster analysis or, data segmentation, …)
Clustering [1]
 Finding similarities between data according to the characteristics
found in the data and grouping similar data objects into clusters
 Similar to classification, except clustering won’t rely upon any
predefined groups.
 Instead the data itself defines the group.
 City-planning: Identifying groups of houses according to their house
type, value, and geographical location
 Earth-quake studies: Observed earth quake epicenters should be
clustered along continent faults
 Land use: Identification of areas of similar land use in an earth
observation database
 Marketing: Help marketers discover distinct groups in their customer
bases, and then use this knowledge to develop targeted marketing
programs
Boise State Executive MBA Program Fall 2015
Han and Kamber, (2006). Data Mining: Concepts and Techniques
 Cluster: a collection of data objects
24
http://userwww.sfsu.edu/art511_h/
acmaster/Project1/project1.html
Clustering [2]
http://today.slac.stanford.edu/feature/2009/data-mining.asp
Boise State Executive MBA Program Fall 2015
http://datamining.xmu.edu.cn/main/~cloud/cbddm.html
25
 Amazon
“People
bought
this also
bought
this”
model
Boise State Executive MBA Program Fall 2015
Han and Kamber, (2006). Data Mining: Concepts and Techniques
Mining
Association
Rules
 Uncovering
relationship
among data.
26
http://timkienthuc.blogspot.com/2012_04_01_archive.html
Data Mining
Life Cycle –
SEMMA [1]
// SAS //
Boise State Executive MBA Program Fall 2015
27
http://timkienthuc.blogspot.com/2012_04_01_archive.html
Data Mining
Life Cycle –
SEMMA [2]
// SAS //
Boise State Executive MBA Program Fall 2015
28
https://the-modeling-agency.com/crisp-dm.pdf
Data Mining
Life Cycle CRISP
Boise State Executive MBA Program Fall 2015
29
https://the-modeling-agency.com/crisp-dm.pdf
Data Mining
Life Cycle –
CRISP Details
Boise State Executive MBA Program Fall 2015
30
Best Practices:
BI Project
Management
Life Cycle:
NOT THIS!
Boise State Executive MBA Program Fall 2015
31
Foundation | SP (2013), What considerations should you have when starting a BI project?
Best Practices:
BI Project
Management
Life Cycle:
NOT THIS
EITHER!
Boise State Executive MBA Program Fall 2015
32
Quartech Systems Ltd. (2013), Business Intelligence... performance from information: Focus your
Organization
Best Practices:
BI Project
Management
Life Cycle:
MORE LIKE
THIS!
Boise State Executive MBA Program Fall 2015
33
Business Value Proposition
of Business Intelligence/
Data Mining
Including Important Best Practices
Boise State Executive MBA Program Fall 2015
34
 http://www.youtube.com/watch?v=THLpOPYj-YY
 Complexity reduction
 Companies on the leading edge of BI are spending 45% less on business
analysis and enjoy 2.4 X on their ROE
 Numerous information silos are being consolidated into a central DW
with Data Marts
 Rensselaer Polytechnic Institute selected Oracle and Hyperion and
noted a payback of 2.5 years, saving them $1.2M.
 Increase competitive advantage
 Now relying upon a single DW providing more timely and accurate
information
sethcurry.ga
How Can
Business
Intelligence
Drive Profits [1]
 The Hackett Group studied 200 large companies in 2007
 Enterprise-wide model construction and minimal number of BI
tools, not desperate siloes and 5-10 tools
Boise State Executive MBA Program Fall 2015
35






 Redefining the role that information
plays in the organization.
 Changing the way that information
requirements are defined.
 Changing behaviors in using
information.
Increase Accessibility
Focus on Accountability
Enhance Accuracy
Timeliness of Information
Improve Decision-Making
Boost Communication
Boise State Executive MBA Program Fall 2015
http://dionhinchcliffe.com/2015/01/16/how-leaders-canaddress-the-challenges-of-digital-transformation/
How Can
Business
Intelligence
Drive Profits [2]
AAATIC (“attic” thread)
Through transformation of
the enterprise culture by:
36
 business strategy :::: BI strategy
 business infrastructure :::: BI infrastructure
 business processes :::: BI processes
How Can
Business
Intelligence
Drive Profits [3]
 Establish a BI Competency Center
 Address new marketplace opportunities:
1. Extending current opportunities. How can firm’s extend
opportunities that are the focus of its current strategy?
2. Potential new marketplace opportunities. What opportunities
beyond the reach of the current strategy should the firm be
considering? What opportunities may be lurking but not yet fully
evident in marketplace change?
Boise State Executive MBA Program Fall 2015
https://whittblog.wordpress.com/2011/04/24/mckinsey-7smodel-a-strategic-assessment-and-alignment-model/
 Establish strategic alignment between
37
 Questioning Competitors
How Can
Business
Intelligence
Drive Profits [4]
 How might competitors most adversely affect the firm’s current
strategy?
 Which competitors are most likely to do so?
 How might the firm best ‘‘handle’’ these threats?
 Questioning Risks
 What competitive risks does the firm’s current strategy face?
 What competitive risks might the firm face in the future?
 How can the firm best manage these risks?
 Questioning the Marketplace Assumptions
 What assumptions about marketplace change underpin the firm’s
current strategy?
 What assumptions should the firm make about emerging and potential
marketplace changes?
 If the firm needs to change its assumptions, what are the implications
for the firm’s strategy change?
Boise State Executive MBA Program Fall 2015
http://leadingstrategicinitiatives.com/2012/04/26/how-to-identify-strategic-assumptions/
 When successful, BI can drive Competitive Intelligence (CI):
38
BI and DMg
Critically Support
Significant
Enterprise
Questions
Potential Value of BI
what happened ?
reporting and dashboards trigger corrective actions, if unfavorable variances are
observed from expectations
why did something happen?
ad hoc queries, user-selected queries, and OLAP are the foundation for
o deeper analysis
o creating customer segments for behavioral analysis
o clustering products that often sell together for marketing purposes
o decomposing enterprise performance information by business unit, product,
customer, or other relevant dimension
why is our enterprise
performance effective/
ineffective?
performance management analytics are intended to enable strategic, tactical, and
operational control over performance and to enable more cost-effective planning
what critical issues/events are
happening right now?
real-time analytics are intended for keeping a tighter rein on minute-by-minute
performance of some key business process
monitoring sales or customer service to enhance such processes, such as by using BI
to suggest book titles to a potential customer who has been looking at other titles
what is likely to happen and
the likely economic results?
predictive analytics are sophisticated statistical and analytical techniques used for
forecasting
Boise State Executive MBA Program Fall 2015
http://leadingstrategicinitiatives.com/2012/11/23/strategicinitiatives-what-are-the-metrics-that-matter/
Questions
39
 Profit margin per transaction
 Margin by customer
 Customer average days to pay
Boise State Executive MBA Program Fall 2015
 Returns count, quantity, and value
 Contribution to profit by product
 Late benefits enrollment
IBM (2005), Business Performance Management Meets Business Intelligence
Best Practices:
Integrate KPIs
 Expiring purchasing contracts
40
http://datamining.typepad.com/data_mining/2008/12/twitter-venn-diagrams.html
Future
Opportunities
for BI, DW and
DMg [1]
Boise State Executive MBA Program Fall 2015
41
http://www.propublica.org/article/world-of-spycraftintelligence-agencies-spied-in-online-games
http://www.entropiaplanets.com/wiki/File:Journal_of_Virtual_Wo
rlds_Research__Payback_of_Mining_Activities_Within_Entropia_Universe.pdf
Future
Opportunities
for BI, DW and
DMg [2]
http://www.cryptocoinsinsider.com
/mining-cryptocurrency/
Boise State Executive MBA Program Fall 2015
42
http://www.worldclassminers.com.au/news/safety/ns
w-miners-build-world-class-virtual-reality-train/
http://livinglabs.mit.edu/index.php?option=com_con
tent&view=article&id=90:original-reality-miningstudy&catid=53:responsive-technology&Itemid=113
Future
Opportunities
for BI, DW and
DMg [3]
http://ostic.wp.tem-tsp.eu/2014/06/17/realitymining-the-technology-which-will-change-our-lives/
Boise State Executive MBA Program Fall 2015
http://news.mit.edu/2013/how-hardit-de-anonymize-cellphone-data
43
http://yellowhammernews.com/nationalpolitics/4th-amend-protectus-warrantless-cell-phone-data-mining-al-06-candidates-sound/
Future
Opportunities
for BI, DW and
DMg [4]
Boise State Executive MBA Program Fall 2015
44
http://www.businessinsider.com/picturesof-the-nsas-utah-data-center-2013-6
Future
Opportunities
for BI, DW and
DMg [5]
--BIG
Challenges
http://blogs.reuters.com/great-debate/2014/01/02/willsnowdens-disclosures-finally-rein-in-the-nsa/
Boise State Executive MBA Program Fall 2015
45
http://www.rippdemup.com/justice/nsa-data-mining-prism-why-myblack-ass-isnt-afraid/
Future
Opportunities
for BI, DW and
DMg [6]
Boise State Executive MBA Program Fall 2015
46
http://timoelliott.com/blog/more-analytics-cartoons
Future
Opportunities
for BI, DW and
DMg [7]
Boise State Executive MBA Program Fall 2015
47
Section 4
Recap
Boise State Executive MBA Program Fall 2015
48