BI to the Masses. End-User BI from a Project Management Perspective. Presented by Dr. Morten Middelfart, TARGIT CTO

Transcription

1 BI to the Masses End-User BI from a Project Management Perspective Presented by Dr. Morten Middelfart, TARGIT CTO

2 Morten Middelfart Work Background 1997 present, CTO of TARGIT A/S, Business Intelligence software company, Denmark , Founder and CEO of Morton Systems ApS, a startup BI software company , BI Project Manager at A.P. Moller/Maersk, Denmark and New Jersey, USA , Platoon Leader, Danish Army, Motorized Combat Infantry , Programmer at Sandlykke & Leifsgaard, Denmark See targit.com/research for more info.

3 Morten Middelfart Educational Background Ph.D. in Computer Science, Aalborg University, Denmark, December 2010 Ph.D. in Computer Aided Leadership & Management, Rushmore University, USA, August 2004 MBA from Henley Management College, UK, February 2002 MISE at A.P. Moller, Denmark, August 1992 Army School of Combat, Denmark, December 1990 See targit.com/research for more info.

4

5

6 The OODA Loop Objective

7 Organizational Context Strategic Tactical Operational

8 Organizational Deployment Strategic Tactical Operational Change Frequency + Diversity = Pluralism

9 Organizational Deployment Operational Tactical Strategic Create / Maintain Equilibrium Change Frequency + Diversity = Pluralism

10 Organizational Deployment Strategic Tactical Create / Maintain Equilibrium Operational

11 Organizational Deployment Strategic Tactical Operational

12 Organizational Context Strategic Tactical Operational

13 Organizational Deployment Strategic Tactical Operational

14 Organizational Deployment Strategic Tactical Operational Change Frequency + Diversity = Pluralism

15 The OODA Loop Objective

16 The CALM Circle (OODA+IT) Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation Data Mining Observation Dashboards Reporting Agents

17 The CALM Circle Decision Search for additional Internal & External Knowledge Search in Un-structured Data KPI(s) Action Storyboards Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle Ad-Hoc Agents & Sentinels Orientation Analysis Simulation Intelligent Analysis Data Mining (Meta-Morphing) Desktop Observation & Mobile Dashboards Notifications Reporting Hyper-Relations Agents

18 The Meta-Morphing Model Used in the TARGIT BI Suite

19 The Cost of a Click Time Risk of Error Education

20 Approaches to Reduce Clicks Intelligent Systems Using Meta-Morphing A Quantitative Approach to User-Friendliness

21 The CALM Circle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation Data Mining Observation Dashboards Reporting Agents

22 The CALM Circle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation TARGIT This & Data Mining Guided Analysis Observation Dashboards Reporting Agents

23 The CALM Circle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation Data Mining Hyper-Related OLAP Observation Dashboards Reporting Agents

24 The CALM Circle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation Data Mining Smiley Observation Dashboards Reporting Agents

25 The CALM Circle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation TARGIT This & Data Mining Guided Analysis Smiley Hyper-Related OLAP Observation Dashboards Reporting Agents

26 What is Meta-Morphing?

27 What is Meta-Morphing? I would like to see customers

28 What is Meta-Morphing? I would like to see customers Show me revenue

29 What is Meta-Morphing? I would like to see customers List my cost and profit per country Show me revenue

30 What is Meta-Morphing?

31

32 Dimensions All Category Product Year Month Day Country City Customers Measures Revenue Cost Profit

33 All Category Product Dimensions Year Month Day Country City Customers I would like to see customers Measures Revenue Cost Profit

34 All Category Product Dimensions Year Month Day Country City Customers I would like to see customers Sentence parse Measures Revenue Cost Profit

35 All Category Product Dimensions Year Month Day Country City Customers Sentence parse customers Measures Revenue Cost Profit

36 Dimensions All Category Product Year Month Day Country City Customers Show me revenue Measures Revenue Cost Profit

37 Dimensions All Category Product Year Month Day Country City Customers Show me revenue Measures Revenue Cost Profit Sentence parse

38 Dimensions All Category Product Year Month Day Country City Customers revenue Measures Revenue Cost Profit Sentence parse

39 Dimensions All Category Product Year Month Day Country City Customers revenue Measures Revenue Cost Profit Sentence parse

40 Dimensions All Category Product Year Month Day Country City Customers Measures Revenue Cost Profit List my cost and profit per country

41 Dimensions All Category Product Year Month Day Country City Customers Measures Revenue Cost Profit List my cost and profit per country Sentence parse

42 All Category Product Dimensions Year Month Day Country City Customers Revenue Measures Cost cost profit country Profit Sentence parse

43 (a) Cube meta-data (b) Questions All Category Product Dimensions Year Month Day Country City Customers I would like to see customers Sentence parse Show me revenue Measures (c) Associations Revenue Cost Profit Sentence parse List my cost and profit per country Sentence parse (d) Presentation preferences Dimension Measure Used Count Customers Revenue 3 Country Profit 2 Country Cost 1 Answer Customers, Revenue Country, Profit Country, Cost Country, Cost, Profit Preference Pie chart Bar chart Bar chart Crosstab

44 The CALM Circle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analysis Simulation TARGIT This & Data Mining Guided Analysis Smiley Hyper-Related OLAP Observation Dashboards Reporting Agents

45 References M. Middelfart. CALM: Computer Aided Leadership & Management. iuniverse, M. Middelfart and T.B. Pedersen. Using Sentinel Technology in the TARGIT BI Suite. PVLDB 3(2): , M. Middelfart. Presentation of data using meta-morphing. United States Patent 7,779,018. Issued August 17th, M. Middelfart. Method and user interface for making a presentation of data using MetaMorphing. United States Patent 7,783,628. Issued August 24th, R.L. Sallam, J. Richardson, J. Hagerty, and B. Hostmann. Magic Quadrant for Business Intelligence Platforms, 2011.

46 Demo and Research can be found at: targit.com/research

47 TARGIT your decisions in fewest clicks Meeting Big Data challenges in Leadership with Human-Computer Synergy. Presented by Dr. Morten Middelfart, CTO

48 TARGIT your decisions in fewest clicks Big Data Data that exists in such large amounts or in such unstructured form that it is difficult to handle in the traditional data warehouse or any other type of database. Right?

49 TARGIT your decisions in fewest clicks From Control to No Control

50 TARGIT your decisions in fewest clicks From Control to No Control Business Intelligence (BI) Reporting, OLAP, and data mining on data that are under our control. Analytics starts where BI ends Typically data will not be under our control.

51 TARGIT your decisions in fewest clicks

52 TARGIT your decisions in fewest clicks From How to Who

53 TARGIT your decisions in fewest clicks But also Why

54 TARGIT your decisions in fewest clicks From Control to No Control No control over availability No control over quality Tipping point for me was 2012 Twenty years of BI, I found myself doing more Analytics of data I did not control

55 TARGIT your decisions in fewest clicks From Control to No Control Do you Fear Change?

56 TARGIT your decisions in fewest clicks What is Fear?

57 TARGIT your decisions in fewest clicks What is Fear?

58 TARGIT your decisions in fewest clicks Strategic Tactical Operational

60 TARGIT your decisions in fewest clicks 14

62 TARGIT your decisions in fewest clicks Facts Beat Fear #HumanComputerSynergy #FactBeatFear

64 TARGIT your decisions in fewest clicks The OODA Loop KPI(s)

65 TARGIT your decisions in fewest clicks The CALM Cycle Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analytics Simulation Data Mining Observation Dashboards Reporting Agents

67 TARGIT your decisions in fewest clicks Any number of unintegrated features or applications will slow you down! - or make you FAIL!

68 TARGIT your decisions in fewest clicks It s a War

69

70 TARGIT your decisions in fewest clicks It s a War Algorithms are already influencing (defining) structured data Scoring the one number analysis Credit Score, Klout -> Profiling Algorithms are moving into unstructured, uncontrolled data too

72 TARGIT your decisions in fewest clicks Synergy between humans

75 TARGIT your decisions in fewest clicks Synergy between humans Ideal

76 TARGIT your decisions in fewest clicks Synergy between humans and computers Ideal

77 TARGIT your decisions in fewest clicks It s a War: Human-Computer Synergy Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Orientation Analytics Simulation Data Mining Observation Dashboards Reporting Agents

78 TARGIT your decisions in fewest clicks It s a War: Human-Computer Synergy Decision Search for additional Internal & External Knowledge Action Storyboard Mass-Comm. Other Communication Workflow Initiation Modify Cycle KPI(s) Ad-Hoc Agents Orientation Analytics Simulation Data Mining Sentinel Mining Desktop Observation & Mobile Dashboards Notifications Reporting Agents

80 TARGIT your decisions in fewest clicks Data Available to Organization M. Middelfart. Improving Business Intelligence Speed and Quality through the OODA Concept. In Proc. of DOLAP, pp , 2007.

81 TARGIT your decisions in fewest clicks M. Middelfart. Improving Business Intelligence Speed and Quality through the OODA Concept. In Proc. of DOLAP, pp , 2007.

82 TARGIT your decisions in fewest clicks Real World Impacts Real World Impacts M. Middelfart. Improving Business Intelligence Speed and Quality through the OODA Concept. In Proc. of DOLAP, pp , Real World Impacts

83 TARGIT your decisions in fewest clicks Real World Impacts Real World Impacts Sentinel Sentinel Sentinel M. Middelfart. Improving Business Intelligence Speed and Quality through the OODA Concept. In Proc. of DOLAP, pp , Real World Impacts

84 TARGIT your decisions in fewest clicks Sentinel Sentinel Sentinel M. Middelfart. Improving Business Intelligence Speed and Quality through the OODA Concept. In Proc. of DOLAP, pp , 2007.

89 TARGIT your decisions in fewest clicks Human-Computer Synergy M. Middelfart and T.B. Pedersen. Using Sentinel Technology in the TARGIT BI Suite. PVLDB 3(2): , 2010.

90 TARGIT your decisions in fewest clicks It s a War Computers can semi- or fully guard our back Primarily on Controlled Data Humans can dedicate themselves more to Analytics on Uncontrolled Data What do (not just my own) customers prefer? Which opinions are trending? Etc.

91 TARGIT your decisions in fewest clicks Big Data Analytics: Trends, Hype, TimeLines

92

93

94

95

96

97

98

99

100 TARGIT your decisions in fewest clicks data is apolitical Marissa Mayer CEO, Yahoo

101

102

103 TARGIT your decisions in fewest clicks The New Frontier: Meat Computing

104 TARGIT your decisions in fewest clicks Meat Computing There is still such a thing as intuition Most companies will be sampling MEAT will make the most distinguishing decisions what to sample or not to sample #HumanComputerSynergy

105 TARGIT your decisions in fewest clicks In Conclusion

107 TARGIT your decisions in fewest clicks Conclusion Business Intelligence (BI) Gives True North through Reporting, OLAP, and data mining. Can be highly automated already today! Analytics starts where BI ends Typically data will not be under our control therefore BIAS RULES!

108 TARGIT your decisions in fewest clicks Beat Fear