What to discuss about data?

Transcription

1 Data Sharing Symposium as a side event of RDA P7 meeting, Tokyo 29 February 2016 Open Science activities with ICSU World Data System, International Enterprise of Long Term Data Preservation Dr. Yasuhiro Murayama Member of Cabinet Office Expert Panel of Open Science Associate member, Science Council of Japan ICSU World Data System Scientific Committee ex officio Natl. Inst. Information & Communications Technology International Programme Office Hosted by Based in Tokyo, Japan 1 What to discuss about data? Data issue Relation with Society, and Mutual Trustworthiness Accelerate Science (and Social Activity) Data Driven Innovation New era of digital information vs. printed records What we need to discuss? We already have TCP/IP, Internet, Web, XML, RDF Is anything else necessary for Information Society & Data Driven Innovation? Issue has multi facets. One is Open Science.

2 Science and Society: research papers and data Conventional Science Method Question Hypothesis Experiment/ Observatoin Analysis/ Discussion Conclusion Data Paper Who does? Immediately? Mandatory? e.g., IPCC report by >3000 scientists policy makers General Society/ political decision making Community consensus Open discussion and criticism Shared in the research community An article is not sufficient for validating results. Reproducibility issue Research integrity issue Data as 1 st class research output Social information asset, provided to the general society Essential for irreproducible natural phenomena Global change, space, living organisms, health... Mutual trust between Science and Society Climate Change Knowledge with Thousands of Scientists IPCC (Intergovernmental Panel on Climate Change) WG1 Physical Science Basis Approx. 1,300 scientists worked for the IPCC WG1. (3,000 4,000 scientists for all WG1 3?) [IPCC, 2013]

3 What to discuss about data? Data issue Relation and Trustworthiness with Society Data Driven Innovation New era of digital information vs. printed records What we need to discuss? We already have technology of Internet, Web, XML, RDF Is anything else necessary for Information Society & Data Driven Innovation? Issue has multi facets. One is Open Science. An analogy: distance from technology to service For Users Application Usage Policy Professional Use Techniques Technology Basis Rainfall Map Weather Radar System Visualization Calibrating estimated rainfall data Radar echo data uncorrected rainfall data Noise reduction, data cleansing Transform: time distance Radar echo from rain Value Added Services Search, analytics Brokering, Aggregation, management infrastructure DOI Metadata, Domain metadata, Data granularity Dynamic data DOI (Digital Object Identifier) Persistent Identifier Data = Digital Object

4 [Kathleen Fontaine, 2015] Data and Science データ引用 被引用度 データ業績 Data citation, credit データマネジメントプラン Data Management Plan (DMP) 研究職採用 昇進評価 Scholar position, evaluation データポリシー Data policy, Data sharing principles メータデータ, 識別子 オントロジ metadata, PIDs, ontology データパブリケーション Data publication Cost Recovery Model Data repository, Long term preservation E インフラストラクチャ E infrastructure, alliance.org/filedepot_download/383/230 RDA Community Capability Model Interest Group Secretary: Univ. of Bath & Microsoft Research Connections

5 Print & Electronic Technologies as Social Info. Infrastructures 百年の印刷文化の基礎支えと 成長途中のディジタル サイエンス Public library (paper media) :8c 351 years Printing press/gutenberg: 1445 First scientific journal: 1665 Intl. Assoc. Academies: 1899 ICSU established: 1931 World Data Center system : 1957 Print Media ENIAC, von Neumann: 1946 Hard Disk Drive: 1956 TCP/IP, dial up (64kbps): 1982 WWW (CERN): 1991 Broadband internet (>1Mbps):~2000 Electronic Media 70 years New global data initiatives: ICSU WDS, RDA etc.:2008~ Toward Global Community of Trusted Science Data Repository/Services for Long Term Preservation Creation of ICSU World Data System on ICSU 29 th General Assembly decision (October 28, 2008): supported by PAST (since 1950 s) WDC (World Data Center) : 50 WDSs at max. FAGS (Federation of Astronomical and Geophysical Data Analysis Services) PRESENT (2008~) ICSU International Scientific Unions data bodies ICSU National Members data bodies ICSU Interdisciplinary Bodies data activities 92 Members (June 2015) 60 Regular Data curation & data analysis services 10 Network Networks of Regular Members & umbrella organizations 4 Partner Do not deal directly with data stewardship, but support to ICSU-WDS 18 Associate Organizations interested in the WDS endeavour 10

6 Opening Ceremony of WDS International Programme Office May 2012) NICT President Prof. Miyahara ICSU President Prof. Y.T.Lee (2012) SCJ former Vice President Prof. Doi Minister of Internal Affairs & Communications Vice Minister of MEXT SCJ President Prof. Ohnishi ICSU (Intl. Council for Science) has established ICSU WDS (October 2008). Science Council of Japan (SCJ) cooperated hosting WDS Intl. Programme Office at NICT, Japan. WDS IPO Executive Director: Dr. Mustapha Mokrane Scientific Committee [Y. Murayama, 2012] [F. Kasuga, 2013] ICSU WDS Objectives Enable universal and equitable access to quality assured multi disciplinary scientific data; Ensure long term data preservation/ stewardship; Foster compliance to agreed upon data standards and conventions; Provide mechanisms to facilitate and improve access to data and data products. In harmony with GEOSS open data policy (ICSU/CODATA contirubted to) 12

7 How datasets are cited by articles (a) Data collection community Physical properties of Hole ## (b) Data sharing community Inequality Article (d) Referential context Health insurance articles Data [Zettsu et al., 2014] National Social Science Survey Reports of the Deep Sea Drilling Project Working Attitudes Population data Income data Pangaea ( 384,815 citations from OAI-PMH Australian Data Archive (ADA) ( Inter-university Consortium for Political and Social Research (ICPSR) ( citations from OAI PMH Dynamics of Infrastructure Edwards, et al Understanding Infrastructure: Dynamics, Tensions, and Design. Infrastructures become ubiquitous, accessible, reliable, and transparent as they mature. Systems Networks Inter-networks system-building, characterized by the deliberate and successful design of technology-based services. technology transfer across domains and locations results in variations on the original design, as well as the emergence of competing systems. Finally, a process of consolidation characterized by gateways that allow dissimilar systems to be linked into networks.

8 オープンサイエンスとデータ出版 引用 data/index.html 15 Research Data Sharing Today Data sharing practices in past Earth science, high energy physics, genomics, Why is it a topic to discuss, in WDS G8 RDA etc.? Some thoughts in context of Open Science Example of past practices Sharing in a specific community or domain: its background culture/knowledge are already shared in the community. Metadata (data attribution, credit, licensing ) can be minimum. Data reuse culture is based on & depends on the community s own norm. Example of discussion in context of Open Science Data: 1st class output of research (cf. research papers) Wish Data be findable, citable, and reusable by anybody in future (like research papers) Info. Management as Information Asset Metadata (data provenance, identifier, etc.) is increasingly important. Information Organization should the key. 16

9 Landscape of Open Science/Research Data Sharing (from my viewpoint) Earth, Space, Physics, Informatics, Seismology Space Sci. Computer Sci. Physics Social Science Linguistics History Psychology WDS Intl. Program Office, Tokyo UNESCO Ocean Data Exchange Ionosphere Space Weather Linguistics etc.(96 Member Bodies) 日本学術会議 Science Council of Japan 2012 Cabinet Office of Japan 総合科学技術 イノベーション会議 Council for Sci. Tech. & Innovation European Open Science Cloud Future Earth (ICSU, UNESCO, UNEP, UNU, Belmont Forum, ) DIAS w/geoss (U Tokyo, JAXA JAMSTEC, NIES etc) 文部科学省 Ministry of Education, Science 科学技術振興機構 Japan Sci. & Tech Agency RDA (Research Data Alliance) 2008 G8 Science Minister Meeting (2013.6) OECD Open Science WG etc. 17 PLEASE KEEP IN MIND: The objective is to promote science and social activity. Any rules or regulations should not discourage scientists/researchers. DON T regulate what we don t yet understand. ( The Data Harvest, RDA Europe, December 2014)

10 12-16 September 2016 in Denver, Colorado, USA FIN.

11 Science is built of facts the way a house is built of bricks (Henri Poincare, 1902) 家がレンガから建てられるように科学は事実から作られる Facts : Data but an accumulation of facts is no more science than a pile of bricks is a house しかしレンガの山が家でないように 事実の積み上げだけでは科学でない Today s Science tends to be advanced, specialized, far from what is sensed in your daily life. Fact as basis of science is increasingly important. Examples of what is going about scientific data in Japan 1957~ 1958~ 1965~ 2006~ 2007~

12 Academic enterprise for long term data preservation ICSU World Data System ICSU 29 th General Assembly decision (October 28, 2008): PAST (since 1950 s) WDC (World Data Center) : 50 WDSs at max. FAGS (Federation of Astronomical and Geophysical Data Analysis Services) PRESENT (2008~) ICSU International Scientific Unions data bodies ICSU National Members data bodies ICSU Interdisciplinary Bodies data activities Opening Ceremony of WDS Intl. Programme Office (May 2012, Tokyo) ICSU President Ministers for ICT, Education & Science Sci. Council Of Japan President WDS IPO Hosted by Based in Tokyo, Japan 23 G Science Ministers Agreement of Open Research Data Open Government Data

13 Open Access to Open Data and Open Science Overview example Database, Repository Self Archiving Data Sharing Institutional Repository [K. Hayashi, 2015] Open Research Data Data journal Full OA (mega) journal (Scholarly activity) Open Access Science Commons Science 2.0+ Open Science (Common activity) Article Access Research Outputs ReUse Research Activity Open Source Open Innovation Citizen Science Creative Commons Open Data (Open gov.) Code for X 2000 s Improve, 2010 s Incremental Redesign, 25 Disruptive? Expert Panel on Open Science based on Global Perspectives (Cabinet Office, Japan) Promoting Open Science in Japan Opening up a new era for the advancement of science Report by the Expert Panel on Open Science, based on Global Perspectives Cabinet Office, Government of Japan March 30, 2015 [H. Manago, 2015] 26

14 Cabinet Office/CSTI:National Principle of Open Science Cabinet Office Expert Panel of Open Science (Dec, 14 March 15) = Final Report was published at the Web site 30 March [H. Manago, 2015] Input to the 5 th National Basic S&T Plan Policy map for Promotion of Open Science [H. Manago, 2015] 04/03/

15 Example of STI Infrastructure in Japan University based Institutional Repository network over the nation For Articles Pilot project for research data started. JAIRO: Japan Institutional Repositories Online Online Cloud JAIRO Cloud service University On premise [K. Yamaji (NII), 2015] Cloud Services for Open Science support for multi-cloud repository hosting theses papers cloud research data cloud Inter-Cloud research logs cloud education data/logs support for procurement cloud providers open science/data open science/data Direct Connect high-speed & secure comm. SINET research data as evidence open education GakuNin Cloud universities research groups class universities 30 [K. Yamaji (NII), 2015]

16 Visit Tokyo 29 February 3 March! Openness is not a final destination, but be interoperable. Fin.

17 wds.org/community/webinars/ webinar 2/RDAWDSPublishingDataIGWebinarIntro.pdf Data Sharing, the Informatics Way: DOI (Digital Object Identifier) for Research Data Data Life Cycle, Its Steps, and Its Stakeholders [Adapted from Hideaki Takeda (2015) ] Library Research Institute Identifiers Resgister Metadata Domain Science Contents Research Project Researchers Produce Produce Produce Preserve Resgister Resgister Publish Revise Revise Revise Identifier s Metadata Domain Metadata Discard

18 Japanese Metadata framework IUGONET (inter univ. upper atmos. obs. network) IUGONET Metadata Database and search system MD Schema Partnership MD Schema Extension For Groundbased Observations EU ESPAS Project Ontology & MD NASA s Metadata Schema SPASE Japan Link Center s Experimental Project for minting/registration of dataset s DOI DOI (Digital Object Identifier) for citing an object digitally. DOI is managed by Intl. DOI Foundation (IDF). DOI can be given only by Registration Agency (RA) under IDF. Japan Link Center (JaLC) is only a RA in Japan. JaLC experiment project: Japan s first attempt to register dataset s DOI Project steering committee: universities, natl. research institutes, & NDL. Japanese Data Centers doi:10.xxxx (DOI prefix) Assign DOI prefix Web Interface Register DOI URL mapping DOI System 9 RAs [Nose et a., 2014] 36