Delivery Systems: the experience of the UK Data Archive

1
Delivery Systems the experience of the UK Data
Archive

• Kevin Schürer
• Director of UKDA, ESDS
• ICA/PRO Conference
• 3 April 2003

2
History

• Archive established in 1968 (as Data Bank)
• Funded by (then) SSRC to provide a service to UK
  HE sector
• Initial focus on government survey data
• Dissemination in early years by punch cards and
  magnetic tapes
• Resource discovery by printed catalogue

3
Current situation (1)

• New distributed service established 1 Jan. 2003
• Economic and Social Data Service (ESDS)
• Still predominately funded to provide service for
  HE/FE sectors
• ESRC, JISC, University of Essex
• Project funding (EC, JISC, MRC, AHRB, etc.)
• Acquire c. 170 studies per annum
• Government 50 academic 30 other 20
• Mixed data types and formats
• Specialist Qualidata unit and History Data
  Service

4
Current situation (2)

• C. 1,750 users per annum
• Servicing c. 5,750 orders per annum
• Distributing c. 14,250 data studies per annum
• Collection equals c. 4,600 studies
• At any one time c. 70 of use concentrates on
  just 20-25 studies

5
Metadata Resource Discovery

• 2 basic levels of metadata
• study level description
• variable level, methodological and technical
  description
• Catalogue searches traditionally at first level
• Catalogue XML database based on DDI
• developed from SSD mapped to other standards
• Several ways to locate studies
• simple searching
• context specific searching
• thesaurus driven searching
• specialist pages

6
Data Delivery

• Several options available depending on user needs
• bespoke writing of CDs and DVDs
• direct web-based download
• web-based online browsing, tabulation, subsetting
  (Nesstar)
• web-based subsetting and download (Census)
• Data availability in various formats

7
Web-based Download

• For bulk Download

8
Nesstar

• EC-funded development work
• Co-project with NSD
• Using DDI and XML structures allows users to
  search and move between both levels of metadata
• Allows for distributed searching retrieval
• Continuous additions to functionality
• aggregate or cube data
• embedding within publications
• geographical interfaces

9
Census subsetting tool

• Designed mainly for larger datasets where users
  will not require entire body of data
• e.g. 1881 census
• Allows slices to be cut out of data according
  to user needs
• Query run offline and resulting dataset placed
  on secure server for download

10
Authentication

• User registration system
• Moving to use of electronic signatures (ATHENS)
• Use of Single-Sign-On system
• Downloading datasets also covered by Access
  Control Units
• Data differentially accessible

11
Delivery and Preservation

• Use of DDI/XML for data delivery
• Can this be used as a long-term preservation
  format?