The document provides an overview of data curation services in academic libraries, emphasizing that data curation is a collaborative effort involving various stakeholders. It outlines the essential components of data curation, distinguishes it from data management, and highlights the development of repository services, specifically the Scholarsphere project at Penn State University. Additionally, it discusses the importance of community engagement and iterative feedback in enhancing data curation practices and services.

1. Data Curation and Data Curation
Services in Academic Libraries -
an Overview
Patricia Hswe | Pennsylvania State University Libraries
Digital Content Strategist
Head, ScholarSphere User Services
CLIR Postdoctoral Fellowship Program Summer Seminar
Bryn Mawr College | 30 July 2014
1

2. My Context
How I Got Here
2
“Cubicle wall” - by sidewalk flying CC BY 2.0

3. First CLIR postdoc cohort
U. Illinois at Urbana-
Champaign: Slavic &
East European Library,
2004-2006
Slavic Digital
Humanities Fellow
Digital projects,
reference services,
workshops Summer 2004
3

4. Library and information
science school - GSLIS
Digital Library track
Document Modeling; Use and
Users; Interfaces to Information
Systems; Digital Libraries;
Information Modeling;
Metadata; Information Transfer
& Collaboration in Science;
Ontology Development
Graduate and research
assistantships - Digital
preservation projects, Engineering
and Mathematics libraries
4

5. National Digital Information &
Infrastructure Preservation
Program
Publishing and Curation Services
Penn State University Libraries
{ Since 2008 }
5

6. Template for the Talk
What is data curation?
Curation is not a solo undertaking.
Developing a repository service - one
piece of the data curation puzzle.
Suggested essentials for getting started.
6

7. What is data
curation?
What is data?
What isn’t data curation?
Purposes.
7
“20080528-002corrupted”- by nvshn CC BY SA 2.0

8. Research data may be -
Observational (e.g., sensor data, data from surveys)
Experimental (e.g., gene sequencing data)
Simulation (e.g., climate modeling data)
Derived or compiled (e.g., text mining, 3D models)
Reference or canonical (e.g., static, peer-reviewed
data sets, likely published and curated)
(from U. of Edinburgh, Information Services)
8

9. Examples of data formats -
Text
Numerical
Multimedia
Models
9
Data specific to a
discipline (e.g.,
Chrystallographic
Information File, or CIF)
Data specific to an
instrument (e.g., images
from a digital
microscope)
from U. Edinburgh, Information Services

10. 10
“Walt Whitman’s Leaves of Grass” by dbking CC BY 2.0

11. 11

12. What isn’t data curation?
Only backing up
data
Not equal to data
management,
though overlaps
Digital preservation
Vs. digital curation
12
“Data” - by UWW ResNET CC BY-NC-SA 2.0

13. 13
Transformation
(migration, derivatives)
Collection of data
(selection and appraisal)
Documentation / description
(metadata)
Ingest / preservation / storage
Dissemination / sharing
access / use / reuse
Standards
Policies,
rights,
conditions
I
N
T
E
R
O
P
E
R
A
B
I
L
I
T
Y
“Gulliver’s collections” by Yohan Creemers CC BY-SA 2.0
INFRASTRUCTURE

14. Examples of data curation
activities -
Organizing data
Reformatting, compressing
(e.g., .tar, .zip), normalizing
Planning for long-term security
Protecting, redacting,
anonymizing data
Applying licenses to data
Providing data citation
guidance
14
“Clean up, or you’re out! :Brooklyn Street Sign” - by
emilydickinsonridesabmx CC BY 2.0

15. Purposes: proper curation of
research data should enable -
Verifiability
Understanding of data use
Identification of responsibility
“Findability”
New research
Improved guidelines, policies
15
“Ecosystem” - by moizissimo CC BY-NC-ND 2.0

16. Keep in mind: use cases
Researchers’ goals
Current research
workflows and
desired workflows
Role / context of
“data keeper”
Intended uses
16
Primary users of the
data
Types of access
Frequency of
access / use
Potential
outcomes / impacts

17. Curation is
not a solo
undertaking
Thinking beyond
models
17
“Soloist” - by Chris JL CC BY-NC-ND 2.0

18. 18
Plan research/
design methodology
Collect/Generate
data
Document/Describe
Process/Analyze
data
Preserve (Prepare for
ingest)
Store data
Share/Access data
Create derivatives /
reuse
A mashup
of a few
models

19. 19
Plan research/
design methodology
Collect/Generate
data
Document/
Describe
MANAGE DATA
Process/Analyze
data
Preserve (Prepare for
ingest)
MANAGE DATA
Store data
MANAGE DATA
Share/Access data
MANAGE DATA
Create derivatives /
reuse
MANAGE DATA

20. 20
Plan research/
design methodology
Collect/Generate
data
Document/
Describe
MANAGE DATA
Process/Analyze
data
Preserve (Prepare for
ingest)
MANAGE DATA
Store data
MANAGE DATA
Share/Access data
MANAGE DATA
Create derivatives /
reuse
MANAGE DATA
C
O
L
L
A
B
O
R
A
T
I
O
N
C
O
L
L
A
B
O
R
A
T
I
O
N
Researchers
Researchers
Researchers
Researchers
Data curators /
librarians, IT
Data curators /
librarians, IT
Data curators /
librarians, IT
Data curators /
librarians, IT

21. 21
Plan research/
design methodology
Collect/Generate
data
Document/
Describe
MANAGE DATA
Process/Analyze
data
Preserve (Prepare for
ingest)
MANAGE DATA
Store data
MANAGE DATA
Share/Access data
MANAGE DATA
Create derivatives /
reuse
MANAGE DATA
C
O
L
L
A
B
O
R
A
T
I
O
N
C
O
L
L
A
B
O
R
A
T
I
O
N
Researchers
Researchers
Researchers
Researchers
Data curators /
librarians, IT
Publishers
Data curators /
librarians, IT
Data curators /
librarians, IT
Data curators /
librarians, IT

22. 22
Plan research/
design methodology
Collect/Generate
data
Document/
Describe
MANAGE DATA
Process/Analyze
data
Preserve (Prepare for
ingest)
MANAGE DATA
Store data
MANAGE DATA
Share/Access data
MANAGE DATA
Create derivatives /
reuse
MANAGE DATA
C
O
L
L
A
B
O
R
A
T
I
O
N
C
O
L
L
A
B
O
R
A
T
I
O
N
Researchers
Researchers
Researchers
Researchers
Data curators /
librarians, IT
Publishers
Data curators /
librarians, IT
Data curators /
librarians, IT
Data curators /
librarians, IT
Users of
the
data!

23. Curation involves building
and sustaining -
Community
Partnerships
Integration
toward
solutions
23
“Relationships” - by beatnic CC BY 2.0

24. Examples of libraries leading
in data curation services
Purdue University Libraries
The Library - UC San Diego
NYPL Labs
University of Minnesota Libraries
24

25. Developing
a Repository
Service
One piece of the
data curation puzzle
25

26. ScholarSphere didn’t
happen overnight.
26

27. Text
Platform review, pilot project, prototype
2010 through 2011
27
Building a Community of Practice: The Curation Architecture Prototype Services
(CAPS) Project at Penn State University
!BACKGROUND:!Review!of!Legacy!Pla:orms!
 !Myriad,!unconnected!workflows!
 !Two!pla6orms!effec9vely!moribund!
 !Another!pla6orm!without!upgrades!since!2007!
 !Impossible!to!search!content!across!pla6orms!
 !No!support!for!eErecords,!liFle!for!research!data!
sets!
 !Lack!of!unifying!architecture!
Co#authors:+Michelle+Belden,+Kevin+Clair,+Daniel+Coughlin,++
Michael+Giarlo,+Patricia+Hswe,+and+Linda+Klimczyk++
CAPS%architecture%–%our%point%
of%departure%
Dashboard%interface%giving%
view%of%objects%in%system%%
Wiki%where%stakeholders%
documented%testing%of%CAPS%
How!Prototyping!a!CuraCon!Service!Led!to!Building!a!Community!of!PracCce!–!The!Story!of!CAPS!
 !Aggressive!Cme!line!–!January!to!March!2011!
 !Key!project!roles:!project!manager,!digital!library!architect,!programmer!analyst,!metadata!
librarian,!access!archivist,!and!digital!collec9ons!curator!
 !Use!cases!kickEstarted!prototyping!
 !Daily!meeCngs!allowed!team!to!check!in!briefly!for!updates!&!discussion!
 !Weekly!mee9ngs!with!stakeholders!(archivists,!subject!experts,!digi9za9on!staff)!!
 !Stakeholders!were!consulted!early!on!for!metadata!needs,!with!concerted!aFen9on!to!
crea9ng!a!framework!allowing!for!annota9on!of!objects!throughout!their!lifecycle!!
 !Followed!agile!methods!to!drive!development,!collabora9on,!feedback,!itera9ons!
 !Listened!to!&!learned!from!stakeholders,!being!as!responsive!as!possible!
 !User!requirements!were!prioriCzed,!but!all!of!them!were!documented;!no!need!was!
dismissed!–!both!to!inform!future!development!but!also!to!assure!stakeholders!their!
perspec9ves!counted!and!were!crucial!to!future!development!and!itera9ons!
 !Survey!of!stakeholders!at!project’s!end!revealed!they!felt!listened!to!by!the!project!team,!
and!the!resul9ng!prototype!reflected!their!requirements!and!deliverables!
 !Next!steps:!chart!a!produc9on!path!and!priori9ze!pilot!projects!for!further!tes9ng!
Prototype%of%public%interface%
for%curated%collections%
Ingest%tool%interface%to%upload%
and%describe%digital%object%
Digital!Library!Technologies!
CAPS Poster - https://scholarsphere.psu.edu/files/r781wj67c

28. In 2012: We did 9 months of
development before beta release . . .
28

29. . . . 9 months during which we
prioritized user engagement
Enlisted liaison librarians
Held bi-weekly “stakeholder”
meetings to discuss use cases
Met with researchers to gauge their
needs, learn their use cases
Demoed the evolving tool, got
feedback
Frequent communications about
progress
Usability testing w/ librarians, faculty
& students; developers present
29
“011 Researcher by Equipment” -
by IPASadelaide CC BY 2.0

30. 30

31. 31

32. 32

33. Uses of ScholarSphere
include -
Supplemental files for electronic theses and dissertations
(ETDs)
Compliance with NSF, NIH, and other funding agency
requirements
“Legacy” data sets - i.e., “I need to get these off my hard
drive” data files
Data sets linked to journal articles (publisher requirement)
Student research (“scholarship as pedagogy”)
Service / platform to help diversify what Libraries collect
33

34. 34

35. 35

36. We aimed for breadth,
rather than depth,
initially.
Because we knew we’d be evolving and
improving the service continually. And we are.
36

37. 37
!
!
!
- 10 releases since
2012
- Next release will be
ScholarSphere 2.0 in
fall 2014: new UI/UX
- Expectation
management is a thing!

38. How engagement is paying off
Inquiries from researchers looking for help with managing
their data >> new partners.
Subject specialist inquiries about the service >> more
faculty and student users.
Inquiries from departments to use service for student
scholarship >> more content.
Feature requests from users >> improved service.
Librarians + researchers understand better the importance
of data & data curation >> opportunities for infrastructure,
collections, publishing, partnerships, outreach, new roles.
38

39. Current work, next steps
Current: ScholarSphere Users Group and
incremental feedback on UI/UX work
Usage statistics
More integration with cloud services
Marketing and promotion / outreach (ongoing)
Next: mediated deposit service, “work” data
model, different metadata templates, citation
support (DOIs), notification framework, Zotero
39

40. Suggested essentials- 1
Find out who your collaborators will be.
Who do you want to collaborate with?
Learn the technology infrastructure at your
library and institution.
Be your own use case.
Learn the basics of project management.
Aim to work in partnership with researchers.
40

41. Suggested essentials - 2
Discover new mentors.
Definitely seek out training that skills you up -
problem-based.
Also: MANTRA, online data mgmt course.
Looking for examples of digital curation in
practice? There’s an active Google group for this.
Cultivate a sense of experimentation & of play by
trying out tools and applications new to you.
41

42. Text
Thank you!
Keep in touch: phswe@psu.edu
42