Expanding Big Data Science: Forward & Backward April 4, 2013Technology Trends, Big Data and...

Expanding Big Data Science:Forward & Backward

April 4, 2013 Technology Trends, Big Data and Data-Driven Decisions

C. Randall (Randy) Howard, Ph.D., PMP Big Data Scientist, Thought Leader, Systems Innovation Analyst, Solutions Architect

Sr. Data Scientist, Novetta SolutionsAdjunct Professor, Mason’s Volgenau School of Engineering

[email protected]://www.crhphdconsulting.net/

May 20, 2014

Expanding Big Data Science: Forward & Backward 2

C. Randall (Randy) Howard, Ph.D., PMP Senior Data Scientist, Novetta Solutions Adjunct Professor, Volgenau School of Engineering, GMU

o Big Data Overviewo Systems Analysis & Design Determining Needs in Big Data o Big Data, Small Details & Time (Metadata)

2013 Teaching Excellence Award Nominee Co-Organizer of Big Data Lecture Series, EIT Award Nominee Member, Data Science Working Groups & Sub-teams

International Author & Speaker 30 years IT & systems engineering, architecture, trouble-shooting,

change & innovation

Ph.D., Information Technology, GMU BS, MS: Information Systems, VCU


Agenda Context: What is Big Data All About?

Forward: Considering Multiple Perspectives

Backward: Refactor/Repurpose Legacy Approaches


Context: What is Big Data Science All About?


Context of Material

How was the big data collected?o Empirical Observations & Applicationso Critical Thinking

Where is it stored?o Case Studieso Feverishly Codifyingo Move from Rescuing to Preventing

What are the results?o Clarifying and Connecting Disparate, Contentious Pieceso Still Working…


My Positions on Big Data

Big Data Scienceo Big Data: Problem & Opportunity Spaceo Data Science: Potential Solution Disciplineo Big Data Science: “Applying Data Science to Big Data”

Technology “Reboot” CAN Usher in New Generation of Capabilitieso Big Data Todayo New “Big Data” Tomorrow

Must Clarify Business Value

Have To Think Horizontally & Corporately

But, I am a professor… Heresy Now? Genius Tomorrow?


IT Disasters & Dilemmas: Possible w/ Big Data? [IT-Failures]

UK Inland Revenue*

$3.5B:Software ErrorsFBI’s Trilogy Virtual

Case File*

$170M:Scrapped

Dis

aste

rs

Dilemmas

Economic Winter(Do more w/ Less)

Ford’s Purchasing System*$400M:Abandoned

NSA Trailblazer *

$1.2B: over-budget, ineffective,

7-yr boondoggle

What is it? Exactly?

Obama Care?


Technology Trigger: A potential technology breakthrough kicks things off. Early proof-of-concept stories and media interest trigger significant publicity. Often no usable products exist and commercial viability is unproven.

Trough of Disillusionment: Interest wanes as experiments and implementations fail to deliver. Producers of the technology shake out or fail. Investments continue only if the surviving providers improve their products to the satisfaction of early adopters.

Peak of Inflated Expectations: Early publicity produces a number of success stories—often accompanied by scores of failures. Some companies take action; many do not.

Slope of Enlightenment: More instances of how the technology can benefit the enterprise start to crystallize and become more widely understood. Second- and third-generation products appear from technology providers. More enterprises fund pilots; conservative companies remain cautious.

Plateau of Productivity: Mainstream adoption starts to take off. Criteria for assessing provider viability are more clearly defined. The technology’s broad market applicability and relevance are clearly paying off.

Curve of Complacency: Early successes satisfy stakeholders that the problem or opportunity is handled, and it is time to move on to the next issue. Meanwhile the Plateau of Productivity that is achieved is much lower.[crh]Dr. C. Randall Howard, PMP (Not a position of Gartner or Dr. Aiken-yet)

[Aiken] [Gartner]

My Big Concern!!


[Conway]

Big Data & Data Science “1-Page Summary”

Big Data “V”s[IBM]:o Volume (How much in total)o Variety (How many sources)o Velocity (How fast does it come in)o Veracity, Variability, Complexity, etc.[various]

“Hard” Data Science[various]

o Math, Science, Analyticso Data-Driven Organizationso Creating data productso Looking to the future

“Soft” Data Science? (Hold on)

Creation & Collection

Capabilities

Time

Data

V’s

Processing & Analytical

Capabilities

Capability gaps due to surges in data collections

NOTIONAL DEPICITION

• Increases in Sensors• Social Media• Mobile Data


Soft Data Science [crh]

Creation & Collection

Capabilities

Time

Data

V’s

Processing & Analytical

Capabilities

Shrink the Capability Gap

w/ “Hard” Data

Science Alone

w/ “Soft” &

“Hard”

Data Science

“Soft Head Start”

NOTIONAL DEPICITION

• Backlogs increase exponentially• Signals become noise• “Action” windows lost / missed• We become bottlenecks to partners

• Notoriety to date• Performed by a few• Bottlenecked by a few?

Changing Term to Tacit Data Science, but that’s another talk

Hardening the “Soft”•Automate “Hard-to-Automate”•Predict Predictable•To-be Performed by Many


Big Data Science Value Parameters

Increased Actionable Intelligence

Trends Noticed / Confirmed

Leverage Unstructured

Faster Knowledge / Awareness / Ability to Search Data

Flexibility / Extensibility of Data Utilization

New, More Adaptable HW/SW Acquisition Models

More TBD


Other Big Data Considerations

Capabilities Their Own Separate ROI’s

Process Data w/in Acceptable Tolerances:o Timeo Errorso Accuracyo Reliabilityo Etc.

Accountability: Find Critical Intelligence & Make Time Windows

Thus, Big Data Is “Having more data than you can process and manage within acceptable tolerances (e.g. time, quality, cost)”[crh]


Forward: Considering Multiple Perspectives


BDLS: A Broader Look Big Data Science

Each channel is difficult

Each complements the other

Complexities are compounded exponentially in cross-sections


Multi-disciplinary[Gartner-ERDS] teams[Patil] a “broad sample of the population” & involves “teams that frequently partner w/ diverse roles in an organization… to gather, organize, & make use of their data”[EMC-DS]

“Wetware[Gleichauf]” (vs. HW & SW): “People, their skillsets, corporate policies, & organizational structures that define our analytic communities”

Soft Skills[Gartner-ERDS]: o Communicationo Collaborationo Leadershipo Creativityo Disciplineo Passion

Data Scientist can be invaluable…unique combination of technical & business skills…makes them difficult to to find or cultivate. [Gartner-ERDS]

Multiple Perspectives in Publications


Data Science Teams

Data Science Teams[Patil]

o Small-team members should sit close to each other o Mix of skill-sets, some experts, some noto Train people to fish o Functional areas must stay in regular contact and communication.

Impedimentso Measuring Performance: Rewarding & Disciplining Teams vs. Individualso Sharing Intellectual Property w/ Integrated Product Teams (esp. cross-vendor)o “Expert Teams”????

“Expert Teams”o May find Big Data Science trivialo Typically

• have more control over their environment• Don’t need to have the masses engaged

But …o Most organizations need to have the knowledge & skills spread out to “Non-experts”


Life-Cycle Service Orchestration

Legal Review

Life Cycle

OODA Loop

Acquisition (FAR)


Classroom Exercise Findings


Wicked Problems


Wicked Problems Tip-off Words[Nixon]

Integrated Joint

Interoperable

Shared

Cross-organizational

Networked

Multi-organizational

Virtual

Coalition

Community

Combined

Big Data is a Wicked Problem!


Wicked Problems[Nixon]

Requires Multiple Stakeholders’ Perspectives Key Driver: Social Complexity from Integrated Networks Traditional linear solution styles are not well suited

Needs focus on:o Social Aspectso Gaining Shared Understandingo Try Thingso Let Solution Emerge From Cycle of Adaptation

Thus[crh], o Multiple Perspectives Involves Collaborationo Collaboration Technologies MUST BE INNOVATED


Sample Collaboration Innovation[InnovationGames]


Sample Collaboration Innovation[InnovationGames]

[InnovationGames] http://innovationgames.com/


Learning Organizations


Learning Organization [Senge]

Peter Senge (http://www.infed.org/thinkers/senge.htm)o Studied how adaptive capabilities developedo The Fifth Discipline(1990) ‘Learning Organization' (LO)

Basic Learning Organization Disciplines:o Systems Thinkingo Personal Masteryo Mental Modelso Building Shared Visiono Team Learning


Learning Organizations’ Disciplines

Discipline Explanation

Systems Thinking

• Cannot understand the parts until you understand the whole[Aiken]

• Balance• Theory w/ Data[Barbara’] • Ideas w/ Tools [Sagan]

System Maps Diagrams that show key elements of systems and how they connect. You may have heard them called Landscape or Ecosystem

Personal Mastery Clarify & deepen our personal vision…of seeing reality objectivelyOR Know yourself

Mental Models

Carry on ‘learningful’ conversations that:•Expose our internal pictures of the world & hold them up to scrutiny•Balance inquiry and advocacy, where people share their thoughts. OR Express Yourself

Building Shared vision

Capacity to hold a share picture of the future we seek to create Has power to encourage experimentation and innovation.

Team Learning Process of aligning & developing capacities of a team to create results its members truly desire


Changing Culture


Culture Obstacles[econBD]


Changing Culture

Examples:o Hard-driveso Management Visibility of Data Processingo Target’s former CEO?

Leadership needs to foster a culture of:o Increased curiosity about datao Rewarding experimentationo Counting “Assists”

Need ‘democratization’, or open-access, of data”[Patil]

o Or Horizontal Orientation / Governance of Data[crh]

Not trivial - Sharing data exposes risks of:o Misinterpretationo Loss of “credit” associated with results from the data


Education


Education

Establish a new baseline of knowledge to advance

Mason’s Big Data Lecture Series Purpose: o Separate Hype from Realityo Have marquee experts expose what in Big Data:

• Is really working and making a difference? • Shows promise?• Has failed? Needs another try? • Are the impediments?

o Convey daunting challenge Is feasible, but still a challenge


Big Data Adoption [IBM-Analytics]


Learning Revolution [Robinson]

Big Data Science is a REVOLUTION that starts (& continues) w/ LEARNINGo Requires new skillso New leadership models

http://www.ted.com/talks/sir_ken_robinson_bring_on_the_revolution.html


Backward: Refactor / Repurpose Legacy Approaches


What is Legacy?

What “brought us here”o Business Basics (e.g., Planning, ROI)o Structured Systems Analysis (e.g. Waterfall methodology, CMMI)

Yes,o Very Cumbersomeo Have Failed tooBut…o Developed by Very Smart Peopleo For Very Similar Issueso Been “Tested”So…..o Re-invent the Wheel?

To leverage:o Consider Context: Intent & Issueso Re-calibrate / Re-factor For Todayo Come Back to “Common Sense”, What Works

Examples: o Meeting Managemento Scaled Agile


Enterprise Architecture

“Process of translating business vision and strategy into effective enterprise change by creating, communicating and improving the key requirements, principles and models that describe the enterprise's future state and enable its evolution.[Gartner-EA]

Short: Simple Structure & Alignment of Technical & Business Capabilities

So…. Take “Business Back to IT”[crh]

Maintain Line-of-Sight to Value[crh]

Focus on the Mission and Mission Capabilities!


Capability Dependencies Hierarchy

Example: Tool x requires staff time for training & learning


Strategic Planning Survey[Bain]

14-year Compilation of:o 11 Surveyso 8,504 respondents

2006: 88% 3.93


Establish Enterprise-wide Decision Criteria

Convey & Carry Commander’s Intent to Execution Levels

Strategy to Tactics Line-of-Sight[crh]


Engineering “Risky Art” Landscape

• Most impactful, hardest to tame, most ignored

• Least concrete, hardest to sell / prove

• Needs the most “innovation attention”


Big Data Lecture SeriesFall 2012

Session 4: Solving the Risk EquationBig Data Systems Analysis & Engineering “So-What”

41

Users

A Big Data Systems Analysis & Engineering “Success” Story

Lots of ways to do this.Lots of requirements.Lots of ways to get requirements across lots of different stakeholders


Wrapup


Big Data Science Postulates[crh]

If Big Data Science is not a technology problem, then let’s focus on the PROBLEM: the non-technology side, or the human-side.

We must perfect the blending of disciplines to educate & train on Big Data Science (vs. perfecting specific disciplines)

Doing what you are doing will not get you out of the fix you are in since it got you in the fix in the first place – innovate and improve!

Our Big Data Science, Analytics & Intelligence is an ENVIRONMENT and a SYSTEM, not an APP


Big Data / Data Science Postulates (cont’d)


How did we do?

One last time…


References


References [1000v] URL: http://www.1000ventures.com/design_elements/selfmade/quaity_cost-4components_6x4.png [Aiken] Dr. Peter Aiken, Data Blueprint, 2012-2013 [arcweb] http://www.arcweb.com/events/arc-orlando-forum/pages/analytics-for-industry.aspx [asq] URL: http://asq.org/learn-about-quality/cost-of-quality/overview/read-more.html [Bain] http://www.bain.com/management_tools/management_tools_and_trends_2007.pdf [Barbara’] Dr. Daniel Barbara’, George Mason University, 2012 Big Data Lecture Series [Batni] Carlo Batini, Cinzia Cappiello, Chiara Francalanci, and Andrea Maurino. 2009. Methodologies for data quality assessment

and improvement. ACM Comput. Surv. 41, 3, Article 16 (July 2009), 52 pages. DOI=10.1145/1541880.1541883 http://doi.acm.org/10.1145/1541880.1541883

[Conway] http://www.drewconway.com/zia/?p=2378 [coq] URL: http://costofquality.org/wp-content/uploads/2011/02/Cost-of-Quality.jpg [crh] Dr. C. Randall Howard, PMP, crhPhDConsulting.net [Crosby] http://www.philipcrosby.com/25years/crosby.html [ct-bdtech] http://cloudtimes.org/2013/06/13/big-data-techniques-for-analyzing-large-data-sets-infographic/ [dddm] http://www.clrn.org/elar/dddm.cfm [DTIC] http://www.dtic.mil/doctrine/new_pubs/ [econBD] http://www.economistinsights.com/analysis/evolving-role-data-decision-making, August 12th 2013 [EMC-DS] http://www.emc.com/collateral/about/news/emc-data-science-study-wp.pdf [Forbes] http://www.forbes.com/sites/christopherfrank/2012/03/25/improving-decision-making-in-the-world-of-big-data/ [FSAM/BAH] http://www.fsam.gov/about-federal-segment-architecture-methodology.php [Gartner-EA] http://www.gartner.com/technology/it-glossary/enterprise-architecture.jsp [Gartner-ERDS] "Emerging Role of the Data Scientist and the Art of Data Science", Gartner, 20 March 2012, ID:G00227058,

Douglas Laney, Lisa Kart [Gartner-HC] http://www.gartner.com/newsroom/id/1763814


References [gayatri-patele-bay] http://www.slideshare.net/AsterData/gayatri-patele-bay [Gleichauf] See Bob Gleichauf’s article: http://www.iqt.org/technology-portfolio/on-our-radar/Big_Data_Advanced_Analytics.pdf [IBM-usingBD] ftp://ftp.software.ibm.com/software/tw/Using_Big_Data_for_Smarter_Decision-Making_v.pdf [IBM]

http://www.ibm.com/developerworks/data/library/dmmag/DMMag_2011_Issue2/BigData/index.html?cmp=dw&cpb=dwinf&ct=dwnew&cr=dwnen&ccy=zz&csr=051211

[IBM-Analytics] http://www-935.ibm.com/services/multimedia/Analytics_The_real_world_use_of_big_data_in_Financial_services_Mai_2013.pdf [Infocus] http://infocus.emc.com/robert_abate/the-business-case-for-big-data-part-1/ Infostory] http://infostory.com/2012/03/28/data-information-knowledge-web/ [InnovationGames] http://innovationgames.com/ [IT-Failures]

o [http://it-project-failures.blogspot.como http://it.slashdot.org/submissiono http://www.sfgate.com]

[Lwanga] The Job of the Information/Data Quality Professional (2010) Lwanga, Walenta, Talburt (IAIDQ Publication) [Madnick] Stuart E. Madnick, Richard Y. Wang, Yang W. Lee, and Hongwei Zhu. 2009. Overview and Framework for Data and Information

Quality Research. J. Data and Information Quality 1, 1, Article 2 (June 2009), 22 pages. DOI=10.1145/1515693.1516680 http://doi.acm.org/10.1145/1515693.1516680

[Mason-BDLS] George Mason University Volgenau School of Engineering Big Data Lecture Series, 2011-2012 [MIT] http://lean.mit.edu/downloads/2010-theses/view-category.html [Nixon] [email protected] - 08/29/2011, Mason Big Data Lecture Series 2011 [Nonaka, Hirotaka, Knowledge-Creating Company] Nonaka, Ikujiro, and Hirotaka Takeuchi. The knowledge-creating company: How Japanese

companies create the dynamics of innovation. Oxford University Press, USA, 1995. [O’Reily] https://docs.google.com/present/view?hl=en_US&id=0AXaXKp9bt6OXZGd4YzlnYmRfNThjMmo4dm5yaA from What is data

science? O'Reilly Radar [p36] http://information-retrieval.info/taipale/papers/p36-popp.pdf [Patil] Patil, D.J., Building Data Science Teams, 2011 [RG] http://www.riskglossary.com/link/risk_metric_and_risk_measure.htm [Robinson] http://www.ted.com/talks/sir_ken_robinson_bring_on_the_revolution.html [Sagan] Dr. Philip Sagan, Infiniti, 2012 Big Data Lecture Series [Senge] http://www.infed.org/thinkers/senge.htm [Talburt] Dr. John Talburt, 2012 Big Data Lecture Series [Tandem] http://www.tandemlabs.com/documents/CPSA2008.pdf


Backup Slides


J. C. R. Lickleider's Man-Computer Symbiosis[Aiken]

Humans Generally Better Machines Generally Better• Sense low level stimuli• Detect stimuli in noisy background• Recognize constant patterns in varying situations• Sense unusual and unexpected events• Remember principles and strategies• Retrieve pertinent details without a priori connection• Draw upon experience and adapt decision to situation• Select alternatives if original approach fails• Reason inductively; generalize from observations• Act in unanticipated emergencies and novel situations• Apply principles to solve varied problems• Make subjective evaluations• Develop new solutions• Concentrate on important tasks when overload occurs• Adapt physical response to changes in situation

• Sense stimuli outside human's range• Count or measure physical quantities• Store quantities of coded information accurately• Monitor prespecified events, especially infrequent• Make rapid and consisted responses to input signals• Recall quantities of detailed information accurately• Retrieve pertinent detailed without a priori connection• Process quantitative data in prespecified ways• Perform repetitive preprogrammed actions reliably• Exert great, highly controlled physical force• Perform several activities simultaneously• Maintain operations under heavy operation load• Maintain performance over extended periods of time

Best approaches combines manual and automated reconciliation!

Expanding Big Data Science: Forward & Backward April 4, 2013Technology Trends, Big Data and...

Documents

Transcript of Expanding Big Data Science: Forward & Backward April 4, 2013Technology Trends, Big Data and...