Kulturarw

1
Kulturarw³

• The Swedish
• WWW
• Archive
• Eller, att fånga den VärldsVidaVäven
• http//kulturarw3.kb.se

2
Background

• Legal deposit, 1661
• Latest revision 1993
• only electronic documents in fixed form
• CD-ROM, diskettes
• New law under way
• secretary from BIBSAM/KB
• Proposal SOU 1998111
• First Swedish web newspaper lost
• printed newspapers 1645
• Kulturarw3 started summer 1996

3
Goals

• All www and gopher pages in Sweden
• pictures, video etc
• .se and generic TLDs
• suecana
• All articles in electronic journals
• All Swedish newsgroups / mailing lists
• Limitations versus RA and ALB

4
Organisation

• Project group two persons
• Steering group four persons
• Reference group representatives from ALB, RA,
  Lund Univ, SUNET etc
• International cooperation
• NWA - Nordic Web Archive
• Nat. Libraries

5
Strategy

• Selection?
• How to know what is important?
• Labour intense
• Collect everything using automatic software
• Gets everything
• Less labour intense
• Computer memory is cheap

6
Strategy

• Take snapshots of the Swedish web a few times a
  year.
• In the future, take newspapers every day, others
  every month etc
• What is Sweden?
• .se
• .com, .org and .net with Swedish
  address/telephone number
• Swedish .nu (Niue)
• Suecana

7
Robot, Software

• Modified version of Nordic Web Indexs robot
  software (NetLab, Lunds univ.)
• Important!indexing is not archiving!
• Save data in MIME format
• Temporary storage, media DLT-tape
• Data rate, 5MB/s
• Capacity, 20 GB
• Durability, 1 000 000 pass
• Data integrity, error detection and correction
• Low cost per GB stored
• Access time too long (?)

8
Statistics

• 15 MURL (including duplicates)
• 240 Gbytes
• 54 000 sites
• 32 300 .se
• 14 800 .com, .org, .net and .edu
• 100 suecana
• 6 800 Niue
• Compare, legal deposit
• Printed materiel 1,7 km/year
• The web approx. 50 km on swedish web.

9
Statistics

• 363 different MIME types found. Many the same,
  some garbage.
• 7.0 M text/html
• 4.2 M image/gif
• 3.0 M image/jpeg
• 0.3 M text/plain
• 0.5 M others
• text/html image/gif image/jpeg text/plain
  comprises 97 of the documents.

10
The Archive

• Goals
• Create copies of the swedish web at several times
  (compare index services)
• Surf the web in space and time
• Search
• Accessible in the future ? migration

11
The Archive

• Disk
• (Optical disk)
• Magnetic tape

HSM Most data on magnetic tapes, staged to disk
when needed
12
The Archive

• What are we archiving?
• Magnetic tapes?
• Bits and bytes?
• Intellectual content?