tansley china

Category: Entertainment

Presentation Description

No description available.


Presentation Transcript

The China Digital Museum Project: 

The China Digital Museum Project Dr Robert Tansley Digital Media Systems Department, HP

HP Labs and China MoE Digital Museum Project: 

HP Labs and China MoE Digital Museum Project Many universities in China have a museum Many objects of historic, cultural, educational and research value Due to space requirements, and size (geographic and population) of the country, museum contents are not optimally utilised Additionally, some physical objects are deteriorating Therefore objects are being digitised Improve access Preserve

Solution Architecture: 

Solution Architecture Build large-scale, distributed digital asset management infrastructure based on DSpace Requires DSpace architectural extensions Inter-DSpace replication Central identifier and object location management Each university museum runs a DSpace instance for management of locally-digitised material Large-scale “data centre” DSpace instances automatically replicate metadata and content from the university museum DSpaces create subject-based “virtual museums” enable centralised preservation activities

‘Virtual Museums’: 

‘Virtual Museums’

Overall Architecture: 

Overall Architecture

Underlying problem: 

Underlying problem Distributed, inter-organisation digital asset management of heterogeneous media Synchronising and replicating metadata and content between DSpace nodes Identifier management; tracking location of multiple copies of objects Distributed security Consistent and uniform access mechanisms


CNRI Handle Protocol


OAI-PMH METS metadata


HTTP GET Content


CNRI Handle Protocol



Technical components built in China Digital Museum project: 

Technical components built in China Digital Museum project DSpace Handle protocol support Export of METS via the OAI-PMH protocol OAI-PMH harvesting tool to retrieve METS and XML metadata about DSpaces from registry Item replication tool Algorithm robust over time to failed replication attempts, corrupt content transfers “Smart” METS ingest tool handles updates without needing to retrieve full object DSpace internal registry of peers Central Registry tool “Show me available copies of X” UI


Applications Centralised, organised by subject (or other) Consolidating the contents of many sites into one Preservation of digital content Multiple copies exist; if one is corrupt, each DSpace can ‘self-repair’ obtaining correct copies from others Sharing stewardship responsibilities Centralise preservation activities


Applications High user demand Many copies available, for use by users located nearby Many “mirrors” of content Sharing and transferring stewardship E.g. Legal deposit requirement for digital material

Possible topologies: China Digital Museums: 

Possible topologies: China Digital Museums

Possible topologies: University Federation: 

Possible topologies: University Federation

Centralised Deposit + preservation, distributed access: 

Centralised Deposit + preservation, distributed access

Future research directions: 

Future research directions Distributed, multiple copy management Lifecycle management of many copies Updates, deletions, preservation format migrations Audits, digital object authenticity Access Control How to ensure DSpace B enforces constraints specified by DSpace A Managing identifiers as first-class objects Access control (who can say they have a copy of X?)

Future research directions: 

Future research directions Replication agreements, “federations” Machine-readable policies, service-level agreements E.g. “I will replicate 2Tb of your content if you do the same for me” Determining responsibilities Service location in a distributed environment with multiple available copies Grid storage integration (e.g. ChinaGrid, SRB) Bitstream replication managed by Grid, not HTTP transfer

authorStream Live Help