Developing Data Analytics Skills in Japan: Status and Challenge
2 likes16,934 views
1) The document discusses the development of data analytics skills in Japan. It notes that Japan lags in producing data analytical talents, with numbers declining by 5.3%. 2) The Ministry of Education, Culture, Sports, Science and Technology (MEXT) started a project to develop talents for big data. The Institute of Statistical Mathematics and University of Tokyo received $130k per year for 3 years to form a network for scalable development of talents. 3) A study found that "datascientists" in Japan have diverse backgrounds and skills spanning business, mathematics, and computer science. While individuals may have data analytics capabilities, it is really an organizational capability. The maturity of organizations acquiring these services is
Developing Data Analytics Skills in Japan: Status and Challenge
- 1. Developing Data Analytics Skills in Japan: Status and Challenge Hiroshi Maruyama The Institute of Statistical Mathematics 7/17, 2014 Hiroshi Maruyama 1 International Workshop on Data Science and Service Research
- 2. 7/17, 2014 Hiroshi Maruyama 2 “Data Scientist: The Sexiest Job of the 21st Century”
- 3. 33/41 7/17, 2014 Hiroshi Maruyama http://www.mckinsey.com/insights/business_technology/big_data_the_next_frontier_for_innovation McKinsey Global Institute: Big data: The next frontier for innovation, competition, and productivity Japan lags in producing data analytical talents
- 4. -5.3% 4/41 7/17, 2014 4Hiroshi Maruyama Japan’s number is even declining …
- 5. MEXT started a project for developing talents for big data 7/17, 2014 5Hiroshi Maruyama ISM + U. Tokyo awarded the grant for three year project Budget: $130K x 3 years
- 6. Goal: To Form A Network for Scalable Development of Talents 7/17, 2014 Hiroshi Maruyama 6 Data Scientists Certific ation Industry Acade mia Share the Vision
- 7. Five Work Streams of the Project ① Communication ② Rotation (internship) ③ Study on Best Practices ④ Develop Course Materials ⑤ Global Linkage 7/17, 2014 7Hiroshi Maruyama
- 8. 7/17, 2014 Hiroshi Maruyama 8 So who are datascientists?
- 9. Mentor Companies INSIGHT DATA SICENCE FELLOWS PROGRAM 97/17, 2014 Hiroshi Maruyama
- 10. 7/17, 2014 Hiroshi Maruyama 10
- 11. “Data Product” example: CouchTube 7/17, 2014 Hiroshi Maruyama 11 “Datascientists” are those who develop working systems with data analytics Scoring based on data analytics CouchTube.net
- 12. “Analyzing the Analyzers – An Introspective Survey of Data Scientists and Their Work” by H. D. Harris, S. P. Murphy and M. Vaisman http://oreilly.com/data/stratareports/analyzing-the-analyzers.csp 7/17, 2014 Hiroshi Maruyama 12 Survey in the US
- 13. O’reilly’s Survey • Web forms (KwikSurveys.com)、5 pages, ave. 10 min. to fill out • Responders： 250 • Skills, experiences, education, self-image, web presence スキルの選択項目（順列） 7/17, 2014 Hiroshi Maruyama 13
- 14. Result of Clustering Non-Negative Matrix Factorization法による 7/17, 2014 Hiroshi Maruyama 14
- 15. Data Scientist Four Types Binita Data Businesspeople • MBA • Consulting • Data analytics manager at a large corporation • Translator between data and executives Chao Data Creatives • Computer science major • Startup company experience • Open source development in spare time • Consider self as a hacker Dmitri Data Developer • Computer Science major • Professional programmer Rebecca Data Researcher • Ph. D. in Science • Originally in academia • Good at writing academic papers but no management experiences 7/17, 2014 Hiroshi Maruyama 15
- 16. In Japan? 7/17, 2014 Hiroshi Maruyama 16
- 17. Study on Current Status • Quantitative： Survey on the applicants for Statistical Skills Certification Test (319 respondents) • Qualitative： Interviews with 20 “DataScientists” – Industry ： Finance, manufacturing, distribution, public sector, IT vendor, consulting firms, … – Size： From freelancers to large – Roles： Analytics in line business, internal consulting, external consulting, 7/17, 2014 Hiroshi Maruyama 17
- 18. Survey contents • Q1-Q3: Demography • Q4-6： Industry, roles • Q7-10: Data analysis works (frequency, purposes, etc.) • Q11-18: Skills – IT/Statistics/Business – and how they learned them • Q19-20: Career path 7/17, 2014 Hiroshi Maruyama 18
- 19. Demography 7/17, 2014 Hiroshi Maruyama 19 Total 319, 11% female
- 20. Q7. Frequency of data analysis 7/17, 2014 Hiroshi Maruyama 20 全くない 月1日 週1日 週2・3日 毎日 0 10 20 30 40 50 60 70 80 90 EverydayOnce a week Once a month 2-3 times a week Never
- 21. On Careers 7/17, 2014 Hiroshi Maruyama 21 A. 全くそう思わない B. 少しはそう思う C. どちらともいえない D. そう思う E. かなりそう思う Q18. Do you think your skills are effectively utilized? Q19. Do you want to have a career as a data analytics professional? Strongly disagree Slightly disagree Slightly agree Strongly agree Neutral
- 22. Q20. Why do you want to be a data analytics professional? 7/17, 2014 Hiroshi Maruyama 22 0 20 40 60 80 100 120 140 160 180 200
- 23. Our clustering result … Established engineer in a large manufacturing company. Does data analytics as a part of line business (e.g., mechanical design, quality assurance, …) Young, eager to be a datascientist, but has little experiences Professional consultant with long experiences in data analytics. Proud of being a data analyst. Female in a SMB company, doing market analysis. Datascientist is an appealing career because of work flexibility. 7/17, 2014 Hiroshi Maruyama 23
- 24. Finding 1： Datascientists have diverse background 7/17, 2014 Hiroshi Maruyama 24 Business school Mathematical Science Commercial science Hard science (e.g., physics, astronomy)
- 25. Finding 2： Data Scientists are “whole mind” skills 7/17, 2014 Hiroshi Maruyama 25 Business Issues Business Decisions ① Find ② Solve ③ Apply Mathematical Formulation Numeric Solution Analyst / modeler True “Datascientist” ISBN-13: 978-4062882187
- 26. Finding 3： Data analytics is a capability of an organization, not of an individual 7/17, 2014 Hiroshi Maruyama 26 VS Datascientist Data Analytics Team
- 27. Finding 4： Maturity of Acquirer's is also important 7/17, 2014 Hiroshi Maruyama 27 Maturity of Acquirers is also important! Statistics Center, President Toya
- 28. Difference between US and Japan 7/17, 2014 Hiroshi Maruyama 28 Data Products Analytics Services Individual Capability Organizational capability
- 29. So What’s Next? 7/17, 2014 Hiroshi Maruyama 29
- 30. 1. Training Programs – Online material – Internship 2. Discussions on Career – Crowd Soucing 3. Acquirer’s Maturity 7/17, 2014 Hiroshi Maruyama 30
- 31. (1) Training: Online Material “Data Scientist Crash Course” 7/17, 2014 Hiroshi Maruyama 31 Contents （20min. × ８） 0. Overview 1. What is Data Scientist 2. Data Analysis 101 3. Visualization and Tools 4. Statistical Modeling and Machine Learning 5. Modeling Time-Series Data 6. Optimization 7. Data Analytics and Decision Making 8. Intellectual Property in Data Analytics
- 32. (1) Training: Internship Program 7/17, 2014 32Hiroshi Maruyama
- 33. (2) Career: Is Freelance Data Scientist a Viable Option? 7/17, 2014 Hiroshi Maruyama 33 Experiment: Post a data analysis task on a crowd sourcing site Igawa, et al., “An Exploratory Study of Data Scientists in Crowd Sourcing,” The 16th Convention of Japan Tele-Work Society, 2014. 10 Workers
- 34. 7/17, 2014 Hiroshi Maruyama 34 Key: How to Distinguish Best Workers? Best Workers Worst Workers Contracted Workers
- 35. 7/17, 2014 Hiroshi Maruyama 35 Best Workers Worst Workers Contracted Workers
- 36. Skill Certification Program is being Developed 7/17, 2014 Hiroshi Maruyama 36 http://www.datascientist.or.jp/
- 37. 7/17, 2014 Hiroshi Maruyama 37 Analytics Skills Service Providing Skills Service Receiving Skills (3) Services: Skills for “Data Analytics as Service”
- 38. “Co-Elevation” in Service Engagements 7/17, 2014 Hiroshi Maruyama 38 Service provider and service receiver both learn from engagements Kijima & Spohrer, 2010
- 39. • Are there skills / techniques / best practices for service providers that facilitate co- elevation during service engagements? – E.g. Some consultants are reluctant to disclose all their knowledge to the client because they fear losing next contracts 7/17, 2014 Hiroshi Maruyama 39
- 40. Thank You 7/17, 2014 40Hiroshi Maruyama maruyama@acm.org Twitter: @maruyama