Technological: Built under the technological constraints of a different time Technological: Built under the technological constraints of a different time - Data storage was very, very expensive
- Data was stored on tapes that had to be read sequentially
Functional: Built for a different purpose - Designed to print cards not for computerized searching or to supply machine-actionable data
Tools to work with MARC Arcane numbering, labeling scheme - Crosswalk to labels?
- Still have to understand meaning and context
- However, the procrustean limits of field numbers and tags limits the flexibility and expandability of MARC
A couple examples in the context of RDA’s aims A couple examples in the context of RDA’s aims
Intended to be atomistic and independent of storage and display Intended to be atomistic and independent of storage and display Many have controlled vocabulary lists Intended to be used by humans and computers - Machine-actionable data is ultimately for the good of people
Example from OLAC experiment: Example from OLAC experiment: Can we teach a computer to identify widescreen and full screen videos?
Full screen Full screen Fullscreen Full-frame Standard format Widescreen Wide screen Letterboxed
250 Edition statement 538 System requirements note 500 Physical description note 505 Contents note (when disc has both versions) 440/830 Series (widescreen) Where the computer stumbles: - 500 General note: Special features … widescreen to fullscreen comparison
300 $a 1 computer optical disc : $b sd., col. ; $c 4 3/4 in. 300 $a 1 computer optical disc : $b sd., col. ; $c 4 3/4 in. 300 $a 1 CD-ROM : $b sound, color ; $c 12 cm
Transcribed information Transcribed information - IMDb example: Laurence Fishburne (as Larry Fishburne)
Stuff that can’t be controlled data, such as summaries and contents note Explanatory notes for controlled data - Stuff that doesn’t quite fit: square pegs, penguins, and ostriches
Not easy to represent in a machine-actionable way
$w in 7xx linking fields is for the control number of the related record $w in 7xx linking fields is for the control number of the related record $0 in controlled fields - 100 $a Bach, Johann Sebastian. $4 aut
- $0 (DE-101c)310008891
- Isn’t clear how it would work if you want to control two elements in one field
Yet another discussion paper at MARBI on Works and Expression in MARC
Humpty Dumpty problem: Humpty Dumpty problem: $8 linking subfield - not widely implemented or used
511 Cyrus Stevens, violin (1st work) ; Pamela Dellal, mezzo-soprano (2nd work)… 511 Cyrus Stevens, violin (1st work) ; Pamela Dellal, mezzo-soprano (2nd work)… 518 The 1st and 3rd works recorded at the Sonic Temple, Roslindale, MA, Dec. 5 and 14, 2001, respectively… 505 Sonata for violin and piano (17:54) -- A packet for Susan (19:59)… 650 Sonatas (Violin and piano) 650 Songs (High voice) with piano… 700 Boykan, Martin. Sonatas, violin, piano. 700 Boykan, Martin. Packet for Susan…
Why do we want atomistic, machine-actionable data and machine-interpretable relationships? Why do we want atomistic, machine-actionable data and machine-interpretable relationships? OLAC prototype as a possibility
Challenges for development With what $? Who?
Jason Thomale’s “Interpreting MARC: Where’s the Bibliographic Data?” Jason Thomale’s “Interpreting MARC: Where’s the Bibliographic Data?” http://journal.code4lib.org/articles/3832 OLAC’s experiment with extracting data from MARC: Code4Lib Journal: http://journal.code4lib.org/articles/775 OLAC report: http://www.olacinc.org/drupal/capc_files/archived_docs/MIW_4.pdf (task force web page: http://www.olacinc.org/drupal/?q=node/27) MARBI discussion papers Controlled Lists of Terms for Carrier Attributes: - http://www.loc.gov/marc/marbi/2011/2011-dp04.html
Identifying Work, Expression, and Manifestation records: http://www.loc.gov/marc/marbi/2011/2011-dp03.html OLAC prototype discovery interface: - http://blazing-sunset-24.heroku.com/
Dostları ilə paylaş: |