Theo Andrew, Edinburgh University Library ... - Theses Alive!

obtainablerabbiΔιαχείριση Δεδομένων

31 Ιαν 2013 (πριν από 4 χρόνια και 7 μήνες)

119 εμφανίσεις

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Choosing Suitable Open Source

Repository Software


Theo Andrew

SPARC IR Workshop 18
th
-
19
th

November

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Talk Outline


Common Open
-
Source Packages


Generic Software Issues


Specific Repository Software Issues


Case study: ERA

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Common Popular Open
-
Source
Packages



DSpace

(http://www.dspace.org/)


MIT, HP, DSpace Federation


EPrints.org

(http://www.eprints.org/)


University of Southampton


Fedora

(http://www.fedora.info/)


University of Virginia, Cornell Univeristy


Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Support & Development



Support from authors



Documentation essential, mailing lists etc.


Continued development



Bug fixes, feature requests, minimal local
development


Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

System Architecture



Modular architecture


Easy to upgrade, develop and customise


Appropriate programming languages


Stable and appropriate database system


PostgreSQL (open
-
source), Oracle (proprietary)


Easy to integrate into current web services


Templates and styles, using language standards (e.g.
HTML/CSS, XML/XSLT)




Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

System Security


Authentication methods


Most importantly: the one you use at your institution, with the
option to insert your own


Authorisation methods


Able to integrate into current institutional information
systems such as staff, student or course lists


Authenticate
-
able content


Provenance metadata, paper
-
trails, data checksums (e.g.
MD5)


Secure supporting systems


Well
-
known, open security systems and coherent standard
architectures


Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

System Administration


Coherent user administration


Different types of user and user groups


Granular, distributable administration


Delegate areas of the system to different administrators


Licensing System


Related to access policies, with separate submitter,
institution and user licences, ideally with a time
-
dependent
facility


Access policies


Possible requirements: domain restrictions, time
-
dependent
restrictions, partial restrictions


Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Record Handling (1)



Metadata Capture


What metadata do you need? Flexible, appropriate schema
(e.g. Qualified DC, ETD
-
MS (E
-
Theses), MARC21)


Customisable Submission System


Collects relevant metadata, and can be modified
conditionally on the fly


Ingest Methods


Standard submission, batch import, harvesting (e.g. OAI
-
PMH (metadata only)), customised insert using native API

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Record Handling (2)



Extract Methods


Native viewing system, batch export, metadata cross
-
walk,
harvest (e.g. OAI
-
PMH (metadata only)), customised extract
using API


Item Wrappers


Multiple files, multiple metadata records/schemas, internal
structure mapping (e.g. METS, DIDL)

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Digital Preservation




Persistent Identifiers


Some available systems: Handle, PURL, URN, DOI, ARK


Migration


On Ingest (migrate submission to open format), or on
request (preserve migration tool)


Viewers


Tools to render the format are preserved


Emulation


The original viewer is emulated in the new system

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Additional Functionality



Coherent data structuring


An internal structure that can represent your institution in
one or more overlaying schemas


Native Browse


Hierarchical browsing, filtering by structure and metadata;
aids indexing by search engines


Native Search


Constrained search locations, using browse functionality to
display results


Full Text Indexing


Public API (Application Programming Interface)


Creating Portal
-
like services within the institution

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

T
heses

A
live

P
lugin for

I
nstitutional

R
epositories

Case Study: ERA

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Development Box


Production Box


Hardware:



Hardware:


-

Dell 1.3GHz PIII



-

Dell Poweredge 1750


-

2x 36Gb HDD


-

3x73Gb RAID 5


-

2x36Gb Mirrored


-

2x73Gb Mirrored

Software:



Software:


-

Red Hat 8



-

Red Hat Enterprise 3


-

Apache 1.3



-

Tomcat 4.1.30


-

Tomcat 4.0.6



-

Java 1.4.2


-

Java 1.4.1



-

Postgres 7.4.2


-

Postgres 7.3.2


-

DSpace 1.1.1


-

DSpace 1.2



-

Tapir 0.3


-

Tapir 0.3

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Metadata Capture

Submit a Thesis or Dissertation

Main Metadata Page

Abstract & Keywords

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

Multi
-
Part Licensing

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software


No systems yet deal with all issues


Some good development work ongoing
with the various packages


Not all issues need to be solved:


To provide an Institutional Repository


For your institution


Choose a package and get going!

Summary

Theo Andrew, Edinburgh University Library

Choosing Suitable Open
-
Source Repository Software

www.era.lib.ed.ac.uk

This work is licensed under the Creative Commons

Attribution
-
NonCommercial
-
ShareAlike License.

Questions?