Data Curation, SPEC Kit 354

Hudson-Vitale, Cynthia; Imker, Heidi; Johnston, Lisa R.; Carlson, Jake; Kozlowski, Wendy; Olendorf, Robert; Stewart, Claire

27 SPEC Kit 354: Data Curation
15. In which of the following ways do researchers deposit data into your data repository? N=46
Both self-deposit and mediated 22 48%
Mediated 14 30%
Self-deposit 10 22%
Other process 0 —
16. Are there individual file size upload limits for your data repository platform? N=45
Yes 35 78%
No 10 22%
If yes, please specify the file size limit. N=35
GB N
0.1 GB 2
0.2 GB 1
0.5 GB 1
1 GB 1
2 GB 22
5 GB 2
50 GB 1
1000 GB 1
Figure 14. Repository platform data file size upload limits

Previous Page Next Page

SPEC Kit 354: Data Curation (May 2017) resources

Free Attachments

Follow Link Data-Curation-SPEC-Kit-354.pdf

Extracted Text (may have errors)

27 SPEC Kit 354: Data Curation
15. In which of the following ways do researchers deposit data into your data repository? N=46
Both self-deposit and mediated 22 48%
Mediated 14 30%
Self-deposit 10 22%
Other process 0 —
16. Are there individual file size upload limits for your data repository platform? N=45
Yes 35 78%
No 10 22%
If yes, please specify the file size limit. N=35
GB N
0.1 GB 2
0.2 GB 1
0.5 GB 1
1 GB 1
2 GB 22
5 GB 2
50 GB 1
1000 GB 1
Figure 14. Repository platform data file size upload limits

Help

26 Survey Results: Survey Questions and Responses
14. What metadata schema are you primarily using for discovery of data? N=43
Dublin Core 30 70%
MODS 9 21%
DDI 8 19%
DataCite 4 9%
Dataverse 3 7%
MARC 1 2%
Other 4 9%
Comments N=43
Citation metadata compiled by Dataverse software.
Consistent with Dataverse 4.0 for general discovery, Dublin Core Terms DataCite 3.1 Dataverse JSON
(for API) used.
DataCite
DD1 — Dataverse MODS — Islandora
DDI (3 responses)
DDI and Dublin Core (2 responses)
DDI, DC, ISO 19115
Dublin Core (18 responses)
Dublin Core and MODS
Geoblacklight, Dublin Core (modified), MODS, MARC, EAD
Local defined schemas, MODS, VRACore4
METS/MODS Dublin Core (Zenodo)
Modified Dublin Core (3 responses)
Modified version of DDI and DataCite within ePrints.
MODS (3 responses)
There are different scopes for discovery (structured metadata exposed for harvest, Linked Data, SEO
for external search engines such as Google, locally indexed metadata for search and browse, etc.) but
the most applicable schemata are qualified Dublin Core and DataCite.
Very basic Dublin Core. Researchers add their own metadata upon deposit. No element is required.
We have a custom developed schema that incorporates elements from other schema but also includes
specific elements that emphasize preservation of data. Our DMS consultants work with researchers to
fill out a document or template that outlines these metadata elements. Once we move to OSF/Fedora,
we will hopefully work with the community to identify a common set of metadata elements that can be
incorporated into OSF for more automated procedures.
We use a custom schema somewhat based on DataCite but expressed in MODS.

Previous Page Next Page

SPEC Kit 354: Data Curation (May 2017) resources

28 Survey Results: Survey Questions and Responses
100 MB per file
200 MB
500 MB through self-deposit mediated deposit doesn’t have an explicit limit although we aim to give
each researcher 2 TB of space.
1 GB per item
2 GB (17 responses)
2 GB for browser-based uploads, no size limit on the backend
2 GB per file for the self-deposit institutional repository with the option to upload more with assistance
from library staff.
2 GB per file per Dataverse policy with option to request more
~5 GB
5 GB unmediated higher limits mediated
50 GB maximum per file
Dataverse: 2GB university digital library: no limit
For self deposit we say 2G, in practice it is whatever.
I believe DSpace has a 2 GB limit we are discussing imposing more stringent limits as we explore
expanding services.
No specified limit from Digital Commons, but there is an unspecified practical limit imposed by the
HTTP protocol used for file uploads.
Self-deposit: 100 MB |system limit =~1TB
Variable
We meet with each researcher to determine the amount of data. Large data sets above 1TB need
special consideration.
SUPPORT FOR EXTERNAL REPOSITORIES
17. Does library staff help researchers prepare or curate their data for deposit to external data
repositories outside of your institution? N=49
Yes 32 65%
No 17 35%
If yes, which external data repositories do you support most often? Check all that apply. N=28
ICPSR 20 71%
Figshare 15 54%
Open Science Framework 14 50%
Dryad 9 32%
Zenodo 4 14%
Harvard’s Dataverse 4 14%
Genbank 2 7%
Other external data repository 5 18%

Previous Page Next Page

Page 1Survey Results click to expand contents

Page 63Representative Documents click to expand contents

Page 64Data Repositories click to expand contents

Page 76Data Curation Services click to expand contents

Page 94Data Curation Infrastructure click to expand contents

Page 100Data Curation Workflows click to expand contents

Page 106Data Models and Metadata Schemas click to expand contents

Page 116Data Deaccessioning Policies click to expand contents

Page 124Data Curation Job Descriptions click to expand contents

Page 131Selected Resources click to expand contents

SPEC Kit 354: Data Curation (May 2017) resources

All section downloads (PDF) click to expand contents

Free Attachments

Extracted Text (may have errors)

Help

loading

SPEC Kit 354: Data Curation (May 2017) resources

SPEC Kit 354: Data Curation (May 2017) resources