Here is the revised (again) tagSet G/M proposal, based on recent discussion.

I hope this addresses most of the concerns that have been expressed. 

Will follow-up with a revised DC mapping, after we get comments on this revised
proposal.



General Principles
------------------
1.          TagSet-G elements are in general assumed to occur within a Z39.50
      record where that record describes some resource -- loosely speaking,
      where the record is a metadata record. Examples are: GILS record, ZDSR
      document descriptor, collection (or object) descriptive record. A
      tagSet-G element is, in general, a metadata element for the described
      resource. This is distinguished from tagSet-M elements that generally
      pertain to (that is, are meta-data elements for) the record itself, or
      pertain to or facilitate the processing of the record (an example of the
      latter is schemaIdentifier).  TagSet-M also includes utility elements
      that may occur as sub-elements of structured tagSet-G elements.
            In the case where the resource described is an integral part of
      the record, tagSet-G elements are used for metadata about that resource.
      In that case, tagSet-M elements are not used as metadata, but may be
      used in the facilitating/processing role described above.  

2.          A schema may reference a tagSet-G element (by tag) using
      a different name, and may qualify its semantics. This in not intended to
      mean that it may arbitrarily assign the element a different meaning. 
            For example, for 'title', TagSet-G would assign semantics such as
      "... common bibliographic usage".  A MARC schema importing the tag might
      assign it AACR2 semantics. The intent of the tagSet-G definition would
      be to limit the semantic scope so that the element would not be used for
      non-bibliographic uses of "title" (for example, to mean "job title", or
      "title to property"; for these, different tags should be assigned). 
            In general, the semantics attributed to an element as it is
      defined in tagSet-G will be weaker than the semantics of that element as
      defined within a schema. The intent is that general elements may be
      inherrited by schema for more specific usage. 

3.    A tagSet-G element may be defined in one of the following two ways:
      (a)   with a universal, flat ASN.1 datatype, e.g. INTEGER, BOOLEAN,
            GeneralString, GeteralizedTime, etc., or some  other
            well-established datatype, for example INtUnit,
            InternationalString. 
      (b)   As a choice between:
            -     a flat, leaf-element, as in (a); or 
            -     a structured element, structured into subelements such that
                  one of the subelements is designated to contain the actual
                  content, and additional elements may serve as "qualifiers",
                  or other utility elements. The "content" element would be
                  the tagSet-M element wellKnown (19). 


TagSet-G Elements
-----------------
Following is the set of proposed tagSet-G elements including the existing set
defined in the standard (the first 19 elements). The new proposed elements are
20 through 35 (16 elements):

title             1
author            2
publicationPlace  3
publicationDate   4
documentId        5
abstract          6
name              7
date              8
bodyOfDisplay     9
organization      10
postalAddress     11
networkAddress    12
eMailAddress      13
phoneNumber       14
faxNumber         15
country           16
description       17
time              18
DocumentContent   19
language          20
subject           21
resourceType      22
city              23
stateOrProvince   24
zipOrPostalCode   25
cost              26
format            27
identifier        28
rights            29
relation          30
publisher         31
contributor       32
source            33
coverage          34
private           35



TagSet-G DataType Definitions
-----------------------------

   Note that 'type's and 'scheme's referred to below are not yet assigned.
   "type" is intended as a qualifier, for example, to qualify 'date' to mean`
   'dateOfLastUpdate'.  "scheme" is intended as a qualifier, for example, to
   qualify 'date' to indicate in what format the date is provided.


1. title
   dataType: InternationalString 
            or 
         structured into following sub-elements:
         -  'wellKnown', tagSet-M element 19, dataType InternationalString
         -  'type' tagSet-M element 23; dataType: Integer or
            InternationalString
         -  'scheme' tagSet-M element 24; dataType: Integer or
            InternationalString

   
2. author
   Same data definition as title.


3. publicationPlace
   dataType: InternationalString
      Question:Should we depricate this and instead adopt a more general
      element, 'place' with qualifier 'type'?

4. publicationDate   
   dataType: InternationalString. 
      Clearly a candidate for deprication, as publicationDate may now be
      covered by 'date' (qualified by 'type').


5. documentId
   dataType: InternationalString. 
      This is a candidate for deprication, as documentId may now be covered by
      'identifier' (qualified by 'type').


6. abstract 
   dataType: internationalString
      This is a candidate for deprication, as 'abstract' may now be covered by
      'descruiption' (qualified by 'type').


7. name
   Same dataType definition as title.


8. date
   Same dataType definition as title.


9.  bodyOfDisplay
10. organization
11. postalAddress 
12. networkAddress
13. eMailAddress  
14. phoneNumber        
15. faxNumber
   For the above (9-11), dataType is InternationalString.

   
16. country 
   Same dataType definition as title, except no 'type' ('scheme' only).


17. description        
      Same dataType definition as title, except no 'scheme' ('type' only).


18. time
   Same dataType definition as title.
   

19. DocumentContent  
   dataType InternationalString.


20. language
   Same dataType definition as title.


21. subject          
   Same dataType definition as title.


22. resourceType     
   Same dataType definition as country.


23. city
24. stateOrProvince  
25. zipOrPostalCode  
   For the above (23-25), dataType is InternationalString.


26. cost       
      dataType:   
         -  InternationalString, or 
         -  IntUnit, or
         -  structured into following sub-elements:
            o  'wellKnown', tagSet-M element 19, dataType InternationalString
               or IntUnit,
            o  'costInfo' tagSet-M element 25; dataType: InternationalString
            o  'costFlag' tagSet-M element 26; dataType: Boolean 


27. format           
   Same dataType definition as country.


28. identifier       
   Same dataType definition as title. 


29. rights           
   Same dataType definition as country.


30. relation           
   Same dataType definition as title. 


31. publisher        
   Same dataType definition as title.


32. contributor
   Same dataType definition as title.

         
33. source           
   Same dataType definition as country.


34. coverage           
   Same dataType definition as title. 


35. private          
   dataType definition defined by schema. 


TagSet-M Proposed Elements
--------------------------
languageOfRecord       22    Same dataType definition as tagSet-G element
                             language

type                   23    INTEGER or InternationalString

Scheme                 24    INTEGER or InternationalString

costInfo               25    InternationalString

costFlag               26    Boolean  ('true' means there is a cost)


Change type definition of tagSet-M element well-known to "Defined by schema;
default InternationalString"



Ray Denenberg
Library of Congress
202-707-5795
[log in to unmask]