cwspace archiving mit opencourseware in dspace dspace federation 2nd user group meeting university...
Post on 24-Jan-2016
217 views
TRANSCRIPT
CWSpaceArchiving MIT OpenCourseWare in DSpace
DSpace Federation 2nd User Group Meeting
University of CambridgeJuly 7 - 8, 2005
v. 20050704_2215
July 8, 2005 CWSpace, an MIT iCampus project 2
DSpace: wide adoption as Institutional Repository
• CONTENT– Scholarly
materials– Research or
Education oriented
• FEATURES– Safe
• professionally archived
• persistent, citable URL
• preserved over archival timeframes
• policies re: removal, etc.
– Findable• search (metadata; full-text)
• browse
• notification e-mails
• disseminated metadata (OAI-PMH)
• Google and DSpace ("Scholar")
AUDIENCEHumans; SysAdmins;
Spiders; Harvesters
July 8, 2005 CWSpace, an MIT iCampus project 3
DSpace: new territory as LOR?Learning Objects Repository
• CONTENT– Educational content:
• Teaching, learning, instruction, assessment
– Courseware materials• Learning Objects
– Compound Digital Objects, Websites…
• FEATURES– Safe
– Findable
– Emphasis on:• Sharing, Re-Use• Aggregation• Evaluation
AUDIENCEHumans; SysAdmins; Spiders ; Harvesters
System-to-System: CLEs; Image gallery tools
July 8, 2005 CWSpace, an MIT iCampus project 4
Project Goal: InterOperability
"To harvest and digitally archive OCW learning objects, and make them available to learning management systems (LMSs) by using Web Services interfaces on top of DSpace."
• re: "harvest" - Yes. OCW has built "Content Exporter”
• re: "archive" - Yes. DSpace has new IMS-CP ingest module; course rendering
• re: "learning objects" - No, but courseware, Yes. Discovered that from OCW will be Courses not LOs
• re: "Web Services"...to
"LMSs" - Yes. Initial version WS simple clients; also to LMS (SloanSpace) prototypehttp://icampus.mit.edu/projects/DSpace.shtml
July 8, 2005 CWSpace, an MIT iCampus project 5
InterOperability: the main goal
July 8, 2005 CWSpace, an MIT iCampus project 6
CWSpace: Standards forPackages & Protocols
• Packages– IMS Content
Package– METS– MPEG21-DIDL– XFDU
• Protocols, APIs– SOAP & WSDL– WebDAV
• JSR 170 JCR
– RESTful (XML, HTTP)
– XML-RPC
July 8, 2005 CWSpace, an MIT iCampus project 7
Work Activity To DateSchematic showing various areas of development: metadata
specifications, Web Services, export and import programming.
July 8, 2005 CWSpace, an MIT iCampus project 8
Metadata for Content Packaging
• METS– Libraries…
• IMS-CP– Education…
• MPEG-21 DIDL– Commercial…
• XFDU– Aerospace…
July 8, 2005 CWSpace, an MIT iCampus project 9
Package Interchange File (PIF)• IMS-CP uses a .ZIP file with a Manifest XML file
and all content files:
http://www.imsglobal.org/content/packaging/cpv1p1p4/imscp_bestv1p1p4.html
July 8, 2005 CWSpace, an MIT iCampus project 10
IMS-CP and OCW Object Model
IMS-CP imsmanifest.xml
July 8, 2005 CWSpace, an MIT iCampus project 11
METS & IMS-CP
Manifests Diagrammed
http://cwspace.mit.edu/docs/ProjectMgt/Reports/SPARC-IR-Workshop/sparc-poster.html
IMS-CP imsmanifest.xml
METS mets.xml
Red arrow: Logical organization to Physical (href)
Blue arrow (METS only): Descriptive metadata is separated from Logical or Physical
Red arrow: Logical organization to Physical (href)
July 8, 2005 CWSpace, an MIT iCampus project 12
XML Elements: METS & IMS-CP
For those who like the pointy angle brackets…
July 8, 2005 CWSpace, an MIT iCampus project 13
METS SIP Profile for DSpace• Work-in-progress
– MIT CSAIL Publications Archive
• Anticipates DSpace 2.0 possible METS as AIP• Preservation and Technical Metadata• Licenses as METS metadata (Deposit; Creative
Commons)• StructMap raises questions for flat DSpace file
storage (Bundles & fileGrps) (Export re-creation)• Option for new PluginManager to manage• DIP work ahead as well; may differ
CWSpaceArchiving MIT OpenCourseWare in DSpace
D E M O (SCREENSHOTS)
July 8, 2005 CWSpace, an MIT iCampus project 15
OCW Course Rendered in Three Systems
OCW………….. DSpace ………….. SloanSpace …………………………….
July 8, 2005 CWSpace, an MIT iCampus project 16
Year One Demo: OCW to DSpace
July 8, 2005 CWSpace, an MIT iCampus project 17
OCW Course Website• Static website• HTML, PDF,
JPG, XLS• Akamai:
Multimedia• LOM XML• Search;
Feedback; Tracking
• Copyright cleared
http://ocw.mit.edu/OcwWeb/Sloan-School-of-Management/15-040Spring2004/CourseHome/index.htm
July 8, 2005 CWSpace, an MIT iCampus project 18
OCW Content Exporter (CE)
• CE generates entire course website (rewriting links)
• CE writes imsmanifest.xml
• CE publishes .ZIP to web page
• CE can publish whole dept.
• CE also used: translation; professors; etc.
July 8, 2005 CWSpace, an MIT iCampus project 19
Client: IMS-CP DSpace Import Driver
• OCW CE not yet WS client
• Simple CGI driver to DSpace session(), auth/auth(), upload(), and ingest() WS
• 4 params (w-i-p)
RESULTS: Success - Created DSpace item 123456789/75
Using collections='123456789/2’
Started session with token='2-1039ad464ee-7665e5a950ff7ff2'
upload response :: <?xml version="1.0" encoding="UTF-8"?>
<uploadservice><URI>http://rotarran.mit.edu:8080/dspace-ws/upload/package-21802.zip</URI><size>8309770</size> </uploadservice>
Ingesting IMSCP package in package-21802.zip
ingest response :: 123456789/75
Session ended"
July 8, 2005 CWSpace, an MIT iCampus project 20
DSpace Item Record for OCW Course
• DSpace info model maps “Item” to OCW “Course”
• Files (all types) are “Bitstreams”
• Metadata: basic LOM-2-DC
• Year 2: Further DSpace dev re: websites, LOs
July 8, 2005 CWSpace, an MIT iCampus project 21
DSpace Serving Full OCW Course• Static
website• HTML, PDF,
JPG, XLS• Akamai:
Multimedia• DSpace
search (Lucene)
• LOM XML to Dublin Core (w-i-p)
• OAI metadata distribution
• Copyright cleared, DSpace, CC licenses
July 8, 2005 CWSpace, an MIT iCampus project 22
Year Two Demo: Other CLE(early preview!)
July 8, 2005 CWSpace, an MIT iCampus project 23
SloanSpace Searching, Uploading OCW CoursePROTOTYPE
• SloanSpace (dotLRN, OpenACS) portal functionality
• Search DSpace (SRW)
• Retrieve via (provisional) DSpace WS
• Directly “Add Course” to dotLRN module for “Learning Object Repository System” (LORS)
July 8, 2005 CWSpace, an MIT iCampus project 24
SloanSpace LORS Navigation for OCW Course
PROTOTYPE
• dotLRN module, for IMS-CP standard
• LOM XML• (Use of
SCORM permits tracking, etc.)
July 8, 2005 CWSpace, an MIT iCampus project 25
SloanSpace LORS Serving Full OCW Course
PROTOTYPE
• HTML, PDF, JPG, XLS
• Akamai: Multimedia
• Lifecycle issues in Year Two work
July 8, 2005 CWSpace, an MIT iCampus project 26
Year Two Deliverables
• OCW Production– Operationalize to scale for OCW archiving (100
courses Fall 2005…)
• InterOperability with other CLE/LMSs– Package: “CWSpace IMS-CP Profile” for SloanSpace,
for Stellar, for Sakai
– Protocol: Web Services use cases, clients for same
• Archived websites– Improved contextual presentation
• Terminology Extraction Tool (CSAIL)– Explore integration into DSpace, OCW
July 8, 2005 CWSpace, an MIT iCampus project 27
CWSpace: Planned Work
• DSpace Platform– Plugin Manager
• Packagers (I/O)• Crosswalks
– Structured Metadata– Stackable
Authentication
• Protocols– Lightweight
Network Interface (LNI)
• SOAP & WSDL• WebDAV
July 8, 2005 CWSpace, an MIT iCampus project 28
Planned Work on CWSpace
July 8, 2005 CWSpace, an MIT iCampus project 29
Existing Interfaces to DSpace
• DSpace Web U/I• SRU/SRW Search• OAI-PMH• Command-line
– DSpace Batch Importer– Media Filters…
http://wiki.dspace.org/NetworkInterfaces
July 8, 2005 CWSpace, an MIT iCampus project 30
Lightweight Network Interface
• New Proposed Interface(s)– DSpace “Web Services”, for CWSpace…
• SOAP & WSDL• WebDAV (Extension to HTTP protocol)
http://wiki.dspace.org/LightweightNetworkInterface
July 8, 2005 CWSpace, an MIT iCampus project 31
A Few (Quick) Thoughts on
Web Services Design
• Standards-based vs. Custom model
• Abstract vs. Strongly modeled specification
• Technology Approaches (SOAP; REST…)
The next few slides provide a (very) brief tour of the topics we are investigating with our Lightweight Network Interface (LNI)
July 8, 2005 CWSpace, an MIT iCampus project 32
Approaches To Expose Your Object Model, 1
• ‘X’ axis: Degree of match to Standards– Custom
• PRO More control for service• CON Harder to interoperate for consumer
– Standards-based• CON Compromises for service• PRO Easier interoperation for consumer
In selecting an API or protocol or approach to exposing your object model to a network interface there are two axes to consider.
July 8, 2005 CWSpace, an MIT iCampus project 33
Approaches To Expose Your Object Model, 2
• ‘Y’ axis: Degree of precision in model abstraction– Abstractly Modeled
• PRO Open to extension, interpretation, wide application• CON You have a specification, but not implementation
guide
– Strongly Modeled• CON Compromises for service; line is drawn in the sand (!)• PRO Implementable directions; line is drawn in the sand (!)
July 8, 2005 CWSpace, an MIT iCampus project 34
Matrix: Expose Object Model
July 8, 2005 CWSpace, an MIT iCampus project 35
Technology Approaches
• SOAP, WSDL– Enterprise developers
– Contains hints re: objects, methods
• RESTful (“Representational State Transfer”)– Developers comfortable with XML markup, HTTP
– Straightforward XML messages over HTTP
• XML-RPC– Early, simplified spin-off from SOAP (ca. 1999)
• WebDAV (Protocol: extension to HTTP)– “Resources” and “Properties” works well with Repository
In then opting for a style of developing the services your Object Model will provide, there are some technology choices to consider.
July 8, 2005 CWSpace, an MIT iCampus project 36
Resources, Properties: WebDAV
DSpace (root) /mylib/dspace-ni/ Community /mylib/dspace-ni/dso_1721.1%2F46 Collection /mylib/dspace-ni/dso_1721.1%2F3549 Item /mylib/dspace-ni/dso_1721.1%2F5543 Bitstream
/mylib/dspace-ni/dso_1721.1%2F5543/bitstream_13 Workflow /mylib/dspace-ni/workflow/wf_23
The server only pays attention to the ''last'' element in a path of DSO's, e.g.
http://myserver/DAV/dso_123456789%2F1/dso_123456789%2F4/dso_123456789%2F13
is the same as
http://myserver/DAV/dso_123456789%2F13
http://wiki.dspace.org/LightweightNetworkInterface
July 8, 2005 CWSpace, an MIT iCampus project 37
Resources, Properties: WebDAV
Collection objectslogo; short_description ; introductory_text ; sidebar; copyright; default_license; provenance
Item objectssubmitter; owning_collection; license; cc_license; cc_license_rdf; DAV:getlastmodified
Bitstream objectsDAV:getcontentlength; DAV:getcontenttype; source; description; format-id; format-description; checksum; checksum-algorithm; sequence-id
July 8, 2005 CWSpace, an MIT iCampus project 38
Example GET URIs
/mylib/dspace-ni/dso_1721.1%2F5543?session=123xyzzy456&package=org.dspace.METS
/mylib/dspace-ni/dso_1721.1%2F5543/bitstream/13?session=123xyzzy456
July 8, 2005 CWSpace, an MIT iCampus project 39
Example PUT URITo add a new item to the collection at handle 1721.1/3549:
PUT /mylib/dspace-ni/dso_1721.1%2F3549?session=123xyzzy456&package=OCW-IMSCP ....package contents...
HTTP/1.1 201 OK Location: /mylib/dspace-ni/dso_1721.1%2F5549 ....other headers....
July 8, 2005 CWSpace, an MIT iCampus project 40
Features Matrix
• Comparing the Interfaces to DSpace with the Features they provide
July 8, 2005 CWSpace, an MIT iCampus project 41
Three Things to Take Awayfrom this talk on the iCampus Project: CWSpace
• We’re working on Packaging Metadata
• We’re working on Web Services
• We’re also working on Archiving Websites
... help us refine the thinking …
July 8, 2005 CWSpace, an MIT iCampus project 42
CWSpaceArchiving MIT OpenCourseWare in DSpace
Thank You. Questions, Comments…
William Reilly, Larry Stone, MacKenzie Smith—MIT Libraries’ Digital Library Research Group (DLRG)
Rob Wolfe—MIT Libraries’ Metadata Services UnitCec d’Oliveira—MIT OpenCourseWare, Technology
http://cwspace.mit.edu