Cataloguing rules for Institutions records

General rules

• official names of institution and department in 110$$a,$$b (native language)
• well-known acronyms in brackets at the end of 110a/b, else in 410
  • example: 110__ $$aEuropean Organization for Nuclear Research (CERN)
  • look at web page, part of url?
• do not touch ICN in 110u, newICNs in 110t to be created according to the rules below
• new records get as well a 110u so that they can be used in spires
  • brief display needs to be changed
• use minimal url in 8564 - sufficient to identify unambiguously the institute
  • “http://www.ast.cam.ac.uk” not “http://www.ast.cam.ac.uk/IoA/IoA.html”
• Institutes at several cities (eg. DESY, Hamburg - Zeuthen)
  • parent inst (possibly without city) plus several child records for core insts
    • if the name of the parent is too general - e.g. "Nat. Tech. U." - the country is appended: "Nat. Tech. U., Argentina"
  • for non-core institutions decide on a case-by-case basis whether to use only parent institution. Ex.: Hitachi
• departments with only few entries in hep (less than 5): use parent inst
  • if a record already exists:
    • gets as newICN 110__t:obsolete
      • record later to be deleted
    • add the ICN (110u) as 110__x to the parent record
      • when the HEP affiliations will be switched to the new system, the ICNs in 110__u and 110__x will be replaced by the newICN in 110__t
• for duplicates follow the same procedure
  • add 110__t:obsolete to the record to be deleted
  • add 110__u of the obsolete record as 110__x to the surviving record
  • add recid of deleted record in 981__a of surviving record
  • example: http://inspirebeta.net/record/910163 http://inspirebeta.net/record/903370
• departments are linked to their parent institution via 510
  • 510__0: recid of parent inst
  • 510__a: newICN of parent inst
  • 510__w: "t"
• cities *use English for ICN and city field 371b, native name in address 371a
  • several 371__b fields if necessary
• beyond ascii
• 667 currently contains very diverse information. If possible find more appropriate place, e.g. 510 (relation to other inst), or historical data in 6781a. If these don’t fit but the info should be public, put it into 680i
• 372a: University (u), Research centre (r), Company (c)
  • need kb for one-letter abbrev

Questions:

• what to do with obscure institutions? create a rudimentary record as soon as an inst appears? Enter in hep without record in inst? define a prefix to identify them?
• 2nd address to go where? another 371? how to display?

Rules to construct ICNs

• use standard abbreviations (see list below)
• only ascii
• use English abbreviations for ICN as far as possible *Allow exceptions if inst is too well known under native name. Ex: Ecole Polytechnique
• ICN: "Institution"[, "city" (if not part of institution name)][, "Department"]
  • use acronym if reasonable (e.g. part of url)
• for a department record never use only the institution name for the ICN
  • The Department of Physics of Cambridge U. should be “Cambridge U., Dept. Phys.” and not just “Cambridge U.” * If the address contains a department the ICN has to as well. * city at the end (after comma) if not part of the name
  • “Royal Marsden Hospital, London” not “London, Royal Marsden Hospital”
• English form of the city
• If no city can be associated with an inst and its name is not unique, the country is appended: "Nat. Tech. U., Argentina"
• don’t use stop words like “of”
• keep the order of words from the official name: University of Cambridge -> U. Cambridge
• if an inst is located in a small place near a large city and is usually associated with that city, use the city’s name in the ICN, although it’s formally not correct
• the same rules for companies

Standard abbreviations for ICNs

see separate topic

HEP cleanup list

de	cleanup 89	tocheck	taken by DESY
ch	cleanup 16	tocheck	done by CERN
fr	cleanup 60	tocheck	taken by CERN
it	cleanup 104	tocheck	taken by CERN
uk	cleanup 48	tocheck	taken by CERN
eu	cleanup 89	tocheck
us	cleanup 239	tocheck
ca	cleanup 29	tocheck	taken by CERN
au	cleanup 10	tocheck	taken by Heath
cn	cleanup 70	tocheck
jp	cleanup 111	tocheck
world	cleanup 239	tocheck

cleanup is a guideline which inst should be looked at
list of core institutes (in order of priority)
1st column: recid
2nd column: ICN
3rd column: total number of papers in HEP

tocheck is a list of oddities
old institutes (no record in HEP since 6 years)
1st column: recid
2nd column: ICN
3rd column: total number of papers in HEP
4th column: dadd - last HEP record
would be nice so check what happend to them

zerohep list of institutes without record in HEP - some are used in HEPNAMES

Precisions for corrections of institutions (Catherine Cart)

See attached file.

• Precisions_to_clean_or_to_catalogue_the_institutions.docx: (Catherine)

• Inst_rules.docx (Kirsten)

List of adresses (DESY)

• dlu.bas.xls:

Variants of adresses as they were converted by DESY. Some are wrong but it gives a good impression what people write as their affiliation. Every institute in INSPIRE has a line of CC | DLU | | ICN recid: Since this is taken from the DESY-site some INST might be missing.

-- AnnetteHoltkamp - 15-Jun-2011