The slides of my talk 'Systems Development & Application / Data Lifecycle Management in King’s Digital Lab', Bodleian Library, University of Oxford, November 30th, 2017.
1. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
2. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Dr. James Smithies
Deputy Director of eResearch
Director of King’s Digital Lab
@jamessmithies
james.smithies@kcl.ac.uk
The Digital Humanities & The Digital Modern (Palgrave Macmillan, 2017)
Systems Development & Application / Data Lifecycle
Management in King’s Digital Lab
3. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Current Research
The Digital Humanities & the Digital Modern (Palgrave Macmillan, 2017).
Digital humanities need to be understood in the context of ‘the digital modern’.
- Reflexive Modernity (Anthony Giddens et al).
We need to develop critical theories that can help us both understand digital media &
culture and build working digital tools / products.
- Postphenomenology (Donald Ihde, Peter-Paul Verbeek et al).
4. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Future Research
The Meaning of the Machine (aka The Epistemology of the Machine)
• An exploration of how humans have used machines, over the long-term, to construct knowledge and meaning.
• Antikythera mechanism / Chinese water clocks >> Digital Humanities / eResearch / High Performance Computing
/ AI.
The IBM Blue Gene/P "Intrepid" supercomputer,
Argonne National Laboratory.
The Antikythera mechanism (Fragment A –
front), 150-100 BC.
Amazon Echo ‘AI’ assistant.
5. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Enabling DH at Scale
@kingsdh
30 years of activity, against a background of rapid innovation and change in Humanities Computing
and Digital Humanities
Centre for Computing in the Humanities (1991) - Prof. Harold Short
Department of Digital Humanities (2011-):
○ ~450 students across 5 Masters and 1 Undergraduate degrees
○ PhD programme
○ 29 staff (16 fte academic; 8.5 fte research only and 4 fte teaching fellows) plus 8
new appointments
6. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Enabling DH at scale
Operational since 2015
@kingsdigitallab
14 staff: Project Manager, Analysts, Software Engineer, Developers, Designers, Systems Manager,
Research Associate
4 VMWare Hosts, 196 VMs, 784GB RAM with ~400G + AWS / Azure
204 digital projects, including 119 inherited: ~5 million digital objects
Business, HR, Operational Plans for 2016–2020 complete; funding, impact & comms strategies
Supported by external funding, under-written internally
8. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
What we do
Development of collaborative research proposals and projects (digital resources, collections, tools,
mobile apps, data analysis, visualization)
Consultancy on topics such as specific digital methods and technologies, writing research grant
applications, or data curation
Sustainability, enhancing and repurposing existing digital resources, research data and tools
Provision and maintenance of a common infrastructure & shared tools for developing digital
resources within the College
Running events, internships, workshops, training and occasional teaching
10% time, innovation and entrepreneurship
9. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Our work ranges from the development of historical databases and digital archives, digital
scholarly editions, map-based tools, data and text analysis, visualization, and eBooks.
Some of our work
10. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Some collaborators and funders
11. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Inherited Projects
5 > old 22
5 < old 77
Total 99
Classics 23
History 23
Digital Humanities 20
English 14
Music 5
Cultural Studies 5
Spanish 4
Other 5
Total 99
Academic disciplines
Age of resource
Java (Tomcat) 26
Python (Django) 52
Wordpress 7
Omeka 9
Typo3 5
Total 99
Platform
Windows (2003) 2
Windows (2008) 9
Linux (Debian 4) 13
Linux (Debian 5) 32
Linux (Debian 6) 33
Linux (Debian 7) 10
Total 99
Operating System
12. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Archiving Process
• Assess: Scholarly value, Impact, Security,
Maintenance cost.1
• Triage: Risk, Funding, PI & KDL availability.2
• Maintain or archive.3
1 ‘Fit’ with FA&H objectives and KDL model (operational and technical
process dimensions).
2 Prioritisation for phases 1 and 2 (immediate need or delay), related
to KDL workflow & resourcing.
3 Integration within KDL systems development lifecycle, or one of ~4
archival approaches.
13. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
After
Pilot phase 29
Phase 1 35
Phase 2 35
Total inherited 99
Service Level Agreement 49
Migration - ITS 10
Migration - other 20
Archived 20
Total 99
Solutions
Triage Results – phase planning
14. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Principles of Systems Development & Application / Data
Lifecycle Management
We’re at an inflection point:
• The DH community needs to be able to illustrate innovation but also continuity
and value for funding £.
• Need to secure past efforts and put in place frameworks that can ensure new
projects are initiated and maintained according to robust management criteria.
• Experimentation as well as institutional responsibility to integrate with Library
& IT, and manage legal and financial risk.
• Scale & complexity of contemporary projects & infrastructure requires
industry-standard methods.
This does, of course, create challenges in the context of ‘pure’ humanities research.
15. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Principles of Systems Development & Application / Data
Lifecycle Management
Some simple solutions:
• We need to develop some broadly accepted assessment procedures, so we
can document community and research impact as well as the potential
(technical, financial, data-related) risks of a DH project
• We should adopt open technical frameworks and standards in order to enable
interoperability of data and long-term maintenance
• We should produce transparent documentation throughout the project’s
lifecycle, to ensure continuity
• DH labs should develop research data management workflows in close
collaboration with Library/ Data center
We need to hide complexity from our non-technical partners, but ensure our
dev processes protect them (and us) from inevitable future archiving and data
management issues.
16. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Software Development Life Cycle
17. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Software Development Life Cycle
18. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Core Development Stack
https://kcl-ddh.github.io/kiln/
Ancient Inscriptions of the Northern
Black Sea
http://iospe.kcl.ac.uk
Askold Ivantchik, CNRS, University of
Bordeaux
Irene Polinskaya, Department of
Classics, King’s College London
19. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Core Development Stack
https://archetype.ink/
http://www.digipal.eu/
Peter Stokes & Stewart Brookes
20. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Core Development Stack
21. KDL Archiving & Sustainability. Bodleian Library. 30th November 2017.
Get in touch