global challenges, enabling cultures: enriching … · global challenges, enabling cultures:...
Post on 17-Aug-2018
221 Views
Preview:
TRANSCRIPT
Global Challenges, Enabling Cultures: Enriching collaborations for the digital age?
Wendy White, University of SouthamptonALPSP Standing on the Digits of Giants, 08 March 2016
5D 360 TB/disc Data Storage
Nanostructured glass and femtosecond laser writing
Thermal stability up to 1,000 degrees Celsius, life of 13.8 billion years at room temperature
Documents recorded using ultrafast laser with 3 layers of nanostructured dots separated by 5 micrometres
Read by optical microscope
http://www.southampton.ac.uk/news/2016/02/5d-data-storage-update.page
2
Presenting UDHR to UNESCO
3
Can we really think very long term?
For preservation as well as
storage?
Not just technical issues
How do we record information so
in 1000 years they know
Organisation – what is the UN?
Concepts – what is a “right”?
Cultural context significant
challenge
Cultural Collaborations: Metadata
Metadata for the longer term will only be collected if it is valued when created which can then enrich future value
4
Collaboration for Research Enhancement by Active Metadata (CREAM)
Goal - “Enable new research to be built on a platform provided by previous, already existing, outputs, by reusing and repurposing, to generate new technology and outputs by “standing on the shoulders of others”
Encourage the use of active metadata across other disciplines by demonstrating the potential for curation, reuse, reproducibility, reinterpretation, and validation.
https://blog.soton.ac.uk/cream/
5Southampton, Edinburgh, University of the Arts, STFC, Nine by Nine
Metadata - how to make it forward thinking?
Produce exemplars
Promote interdisciplinary reuse
Make use of existing standards
Active metadata needs to be part of active researcher’s activity 6
Procedural Blending Model (Garrelfs,
2015) – facilitate recording of decision
points and motivations as well as
provenance, objects etc.
Can help bridge link between mapping
concepts and process
iterative, agile, creative
Allotrope Foundation
7
• Subject Matter Experts• Project Funding
Member Companies
• Project Management• Legal & Logistical Support
Secretariat
• Framework Development• Technical Leadership
ProfessionalSoftware Firm
• Requirements & Specifications• Contributions, PoC Applications
Partner Network
AbbVieAmgenBaxterBayer
Biogen
BoehringerIngelheim
Bristol-Myers Squibb
Eli Lilly
Genentech/Roche
GlaxoSmithKline
Merck & Co.Pfizer
ACD/Labs
Agilent
Biovia
BSSN
IDBS
Mestrelab Research
Mettler Toledo
Persistent
Riffyn
Sartorius
Shimadzu
Thermo Scientific
Waters
Erasmus Univ. Med
CenterUniversity of Southampton
@2015 Allotrope Foundation
PIs as change agents
• Academic lead, library co-ordination and/or input
e.g. ORCID take up
Academic buy-in
Tech development behaviour led
PIs are editors
Library & publishers working actively & closely with PIs in scholarly communications environment
Principles for collaboration
• Work within framework of existing standards and identifiers e.g. ISO, NISO, WC3, OAI, OASIS, ORCID
• Enabling platform neutral take up of standards and data exchange
• Collaboration must have active research input plus specialist input (libraries, software developers, project management etc.)
• Stakeholders with different expertise valuable – core principles help ensure benefits not stasis due to differing views
10
Research Facilities and Equipment Sharing
DATA SHARING LANDSCAPE IN ACADEMIA
Academia
data.ac.uk data sharing and development community
equipment.data.ac.uk National equipment sharing web
portal
UNIQUIP
Project
(Network,
standards and
outcomes)
N8
GW4
Project deliverables feeding
into data.ac.uk
Accessing “data”
know how
Contributing to or
searching equipment data
M5
SES5
RCUK
Strategy & policy
drivers
Sector technology and
standards e.g. ePrints,
Pure, Agresso, Kit
Catalogue, CERIF,
CASRAI and ORCID
Data and
technology
standards sharing
(creating added
value datasets)
RCUK
“Gateway to
Research”
Project
Sharing knowledge and
outputs with industry
providers (e.g. user groups
and knowledge networks)
Regional consortia
projects contribution
to project
JISC
Support for sector
technology and standards
development
Input from development
projects into Gateway to
Research portal (e.g. G4HE,
CERIF in Action, CASRAI UK
Dictionary)
Knowledge exchange
through UNIQUIP
network
Creating visibility
of research
outputs
Diagram - Adrian Cox
Cultural Collaborations: Publishing
Published data is only useful when it is fully reusable and innovative data visualisation drives engagement
11
Lab Notebook granular data visualisation
12Simon Coles, Andy Milstead
Visualise live changing data, DOIs for snapshots
Quick wins for all
• Embedding a DOI for the data in the publication (could be in a book not just articles) adds value for everyone wherever the data is hosted and stored
promotes data creators, authors, publishers & institutions
• Data host to provide CCO metadata and easy assignation of CC, Government or other reasonable reuse licence
• Helps to recognise:
• More than one provider may have a good reason for needing to incorporate the metadata and possibly dataset in a repository
• Different environment to articles as institutions usually hold the rights to the data not the data creators
• Institutions as well as individual researchers are both consumers and contributors to publishing
13
Learned Societies and Publishers as Innovators
• Zika initiative positive, but should not be exception
http://www.wellcome.ac.uk/News/Media-office/Press-releases/2016/WTP060169.htm
• Can engage with new publishing models and initiatives
Example: Greynet www.greynet.org
Link grey literature and data
Working with national Data Archiving Services and Open Preservation Foundation
14
Preparing and depositing data – journal
15
16
Networks of repositories
• National and international disciplinary centres
• Institutional repositories - increasingly developing to improve preservation
metadata quality, file formats JHOVE, integration with services like Arkvium, archivmatica
• Data journals
• New UK National initiatives – Jisc Research Data Discovery Service, Research Data Management Shared Service
17
All of these types of repository can contribute at a useful
stage in a researcher’s workflow depending on context
and policy environment
Unlocking theses data
• Survey of institutions and case studies looking at current practice and ideas for future practice
• Aim for workflows which capture data to promote reuse as well as thesis text
• Relationships between objects and identifiers - DOIs and ORCID
• Aligned with British Library for integrated national and institutional services
http://unlockingthesisdata.wordpress.com
Other great Jisc dataspring projects http://researchatrisk.ideascale.com/
18with wide community
input in UK
Shared Service Engagement
EnthusiastCommunity
Engagement
Organisational Commitment
19
Shared Services
Further Innovation Communities of Practice
Informal groups
Research Data Alliance
RLUK, ALPSP
Standards/Profiles
CASRAI, OPDs
Aggregated Services
UK Research Data
Discovery
Archival Services
Cultural Collaborations:Learning
Successful data preservation is dependent on developing a culture of continuous reflective learning and practice where researchers, librarians, technicians, publishers & data scientists co-develop networks of knowledge transfer & support 20
Library triaged research data services
Layered approach to networks
• Doctoral College workshops
• Academic input through disciplinary /interdisciplinary modules and bespoke courses
• 1-2-1 deskside consultancy service
• one stop shop advisory service
• collaborations with publishers and editors
• Specialist partnership advice e.g. Software Sustainability Institute 21
22
PICK & MIX
INTENSE &
SEASONAL
EMERGENCY BOOST
INTEGRATED PATHWAYS
Reviewing curricula
Content People
Mode Time
23
Module & course profiles
Level
Themes
PGR Directors
Programme & Module Leaders
Supervisors
PGR & ECR researchers
Workshops
Labs
Online
Point of need
Progression
pathway
Modular
Common data focussed PGT & PGR modules
ETHICS
DATA ANALYSIS
RESEARCH METHODS
24
Working collaboratively with PhD and Early Career Researchers: agents for change
DataPoolBuilding Capacity, Developing Skills, Supporting Researchers
http://datapool.soton.ac.uk/
Project key aims and activity: • Working to build good research data management practices across all disciplines
throughout the data lifecycle, focusing on cultural change and embedding activity.
• Implementation of a formal Research Data Management Policy with associated one-stop-shop guidance and desk-side support for Data Management Planning
• Developing formal and informal support for research data management
• Evolve technical improvement improvements to institutional systems to enhance management, storage and accessibility
Imaging Case StudiesSupporting PhD and Early Career Researchers to investigate 2D and 3D imaging requirements across disciplines“The Datapool 3D and 2D Raster case studies represent a chance to find out how specific kinds of data are created, used and managed across a large institution such as the University of Southampton. The Datapool project recognises that effective data management policy stems not only from ensuring the efficient use of facilities but also from developing deep understandings of how we work and the role which data plays in our research. This aspect of the project aims to survey facilities and resources dedicated to working of these 3D and 2D Raster data and to find out how these resources are used. Through making contact and interviewing individuals and groups we are developing better understandings of how different communities work and collaborate and manage their use of digital media.” Gareth Beale and Hembo Pagi
MultidisciplinaryWorking with events through the existing network of University Strategic Research Groups which focus on multidisciplinary approaches to tackling leading edge societal issues, including the Southampton Multidisciplinary Research Forum for Early Career ResearchersFor example, Twitter Harvesting using ePrints workshop with Web Science DTC reported on Digital Economy USRG blog
Training modelDeveloping training for PhD students with a Service/PhD student co-delivery model for Faculties and an organic model of exploring issues with the Web Science Doctoral Training Centre through their seminar series and spin-out activities
Training Materials
Case study based training guide"We looked at five researchers' work from medicine, materials engineering, aerodynamics, chemistry and archaeology, and produced case studies showing the similarities and differences between the data types they produce. A guide for first year postgraduate students was created containing the case studies and an introduction to research data management. The concepts in the guide has been presented as training lectures, ensuring students start considering the problems associated with research data management early in their careers. The feedback from students suggest that being made to think about these issues is necessary and useful, and engaging them at this stage helps cultivate good practices.“ Mark Scott
Future: Build on structure by adding case studies in other disciplines, for example, English
Technical testingA group of PhD researchers from Music, Physics, Medicine, Geography and Archaeology engaged with user testing for data deposit and management. First test was the SharePoint development with a mixture of structured feedback through a template and informal iteration and mapping to their research questions. Lots of discussion and lunch
Embedding and working with other data management developments to aid support for all disciplines
For example,
“The Heterogeneous Data Centre (HDC) project (JISC funded, Materials Data Centre) is delivering a system to encourage sharing of data sets between materials engineering and medical communities. It was built for managing materials engineering data ranging from small files (kilobytes) to very large (gigabytes, for example, microfocus computer tomography data), using a file system for file storage and a monitoring service to update a metadata database when data sets change. An interface for managing additional data set features has been written in Microsoft SharePoint. The system supports data sets, data set metadata, relationships and collections of data sets, security (to grant others access to a data set), plugins at the data set and file levels, search, data set recommendations, and compatibility with Eprints - the main data catalogue being developed through the DataPool project. The generic approach we have adopted is enabling us to use it in a wider range of application areas such as Medicine.” Mark Scott
Acknowledgements:Authors: Byatt, D., Beale, G., Earle, G., Pagi, H., Scott, M., Coles, S., White, W.
References: Scott, M., Boardman, R., Reed, P., & Cox, S. (2012) Introducing Research Data Faculty of Engineering and the Environment, University of Southampton
(http://eprints.soton.ac.uk/338816/ )
Future – “Across all disciplines”Further expand support across all disciplines - ? Embed digital data experts in disciplines as nodes for multidisciplinary knowledge transferAgents for change - Informing and piloting new developments, developing new case studies
Scott, M et al (2012)Medical Data Fig. 5 & 6 Introducing Research Data. University of
Southampton p.9
Scott, M. et al (2012) Data lifecycle Fig.1 Introducing Research Data. University of Southampton p.4
http://digitaleconomy.soton.ac.uk/
http://tweets.soton.ac.uk/
Reflectance Transformation Imaging capture of brickstamps in Italy. Portus
Project Photo: Hembo Pagi, 2011
Rock Art Libya Photo: Hembo Pagi, 2009
Portus Project 3D laser scanning: Researchers are producing more and more 3D data within increasingly diverse research
contexts. http://www.portusproject.org/Photo: Gareth Beale, 2012
Portus Project 3D laser scanning: Researchers are
producing more and more 3D data within increasingly diverse
research contexts. http://www.portusproject.org/
Photo: Gareth Beale, 2012
http://www.southampton.ac.uk/library/research/researchdata/
26
Skin health sensing Data management
Literature (636 documents)
Clinical (36)
circulation (39)
ulceration (14)
Infection diagnosis (1)glycemic control ( 15)
Gait analysis
Sweat (5)
neuropathy (19)
costs (7)
Measurableperameters
Mechanicalproperties of skin (41)
humidity (2)
temperature (6)
Bioimpedance (75)
oximetery/perfusion (21)
Activety (3)
AccelerationForce/Force (35)
PH (3)
Fatigue (3)Shear (22)
Ultasonic (10)
background (48)
Technology (88)
Data monitoiring (21)
Orthotics/prosthetics (20)
Testing (48 formal)
Ethics approval (10)
Test protocols (5 and increasing)
Test data
Calibration protocols
Calibration data
Outputs
Thesis
9 month report (1)
Transfer report (1)
Final Thesis (1)
Papers (some)
Poster (4)
Presentaions (5)
Sensor design
Sensor selection (9)
Sensorexcitaioncircuit (9)
Sensor calibration (xx)
Sensor validation (xx)
Sensor software (19)
Protoytypesoftware (>30)
Dataaquisition
software (3)
Prototype software (10)
Sensor prototypes (15)
Sensor layout (5)
Design drawings (9)
Meeting notes (53)
Notebooks (40)
Onenote (13)
Training/lecture courses (7)
Notes
Confermation ofattendance
Results
Shared vision and capacity building
27
The effective use of Data Science technologies requires new skills and demands for new professions, usually referred as the Data Scientist: an expert who is capable both to extract meaningful value from the data collected and also manage the whole lifecycle of Data, including supporting Scientific Data e-Infrastructures.
The future Data Scientists must posses knowledge (and obtain competencies and skills) in data mining and analytics, information visualisation and communication, as well as in statistics, engineering and computer science, and acquire experiences in the specific research or industry domain of their future work and specialisation. We call this profession the Data Science Professional (DSP).
http://edison-project.eu/news
28
Embedded
Librarianshttp://www.itutility.ac.uk/
Research Data - more than management
• Need to integrate support with areas of active research interest
• RDM often implicit in ethics/analysis/methods modules - could do more to foreground this and integrate
• Support for more “full data journey” stories from ethics to sharing – need to make more links between ethics processes, data management planning, data capture & data analysis, data publication, data preservation
29
top related