Page Comparison

Subject

Yes

Yes

Yes

No

No

No

No

No

Author

Yes

Yes

Yes

No

No

No

No

No

Format

Yes

Yes

Yes

No

No

No

No

No

Language

Yes

Yes

Yes

No

No

No

No

No

Publication Date

Yes

Yes

Yes

No

No

No

No

No

Genre

Yes

Yes

Yes

No

No

No

No

No

ISSN

022 - a,z (MV)

first value

all values

same as search field

ISBN

020 - a,z (MV)

first value

all values

same as search field

Author/Creator

For each 100, 110: every subf except $6 (gives us 2 values for every tag). Also every subf except $t and $6 for: 111, 700, 710, 711, 800, 810, 811, 400, 410, 411) 
Ref: http://www.loc.gov/marc/bibliographic/
100 - Main Entry - Personal Name (NR)
110 - Main Entry - Corporate Name (NR)
111 - Main Entry - Meeting Name (NR)
700 - Added Entry - Personal Name (R)
710 - Added Entry - Corporate Name (R)
711 - Added Entry - Meeting Name (R)
711 800 - Series Added Entry -Meeting Personal Name (R)
800 810 - Series Added Entry -Personal  Corporate Name (R)
811 - Series Added Entry -Meeting Name (R)
400 - Series Statement/Added Entry - Personal Name(R)
410 - Series Statement/Added Entry - Corporate Name(R)
411 - Series Statement/Added Entry - Meeting Name(R)

first non-empty value of
100$a or 110$a.
Show blank if both are
missing.

All non-empty
indexed values

All non-empty
indexed values

Title

245 - all subf exc.

All non-empty
indexed values

All non-empty
indexed values

Title

245 - all subf exc. c and 6. Also, 130, 240, 246, 247, 440, 490, 730, 740, 773, 774, 780, 785, 830, 840) (MV)

245$a and 245$b

all values

Place of Publication

260 - a (MV)

first value

all values

same as search field

Description

505 - a (MV)
KG/LR: UPenn just included the MARC 505 in its Description index (which is distinct from its Format/Description index). Include just 505 $a.  The SMEs may want additional 5xx fields in the Description index, but 505 should be fine for November.

first value

all values

same as search field

Subject

600, 610, 611, 630, c and 6. Also, 130, 240, 246, 247, 440, 490, 730, 740, 773, 774, 780, 785, 830, 840) (MV)

245$a and 245$b

all values

Place of Publication

260 - a (MV)

first value

all values

same as search field

Description

505 - a (MV)
KG/LR: UPenn just included the MARC 505 in its Description index (which is distinct from its Format/Description index). Include just 505 $a.  The SMEs may want additional 5xx fields in the Description index, but 505 should be fine for November.

first value

all values

same as search field

Subject

600, 610, 611, 630, 650, 651, 653, 69X: every subf exc. $6, $2, $=, $? across these tags

600 - Subject Added Entry - Personal Name (R)
610 - Subject Added Entry - Corporate Name (R)
611 - Subject Added Entry - Meeting Name (R)
630 - Subject Added Entry - Uniform Title (R)
650 - Subject - Added Entry -  Topical Term (R)
651 - Subject Added Entry - Geographic  Name (R)
Ref : http://www.loc.gov/marc/bibliographic/bd6xx.html

first non-empty value of 600, 610, 611, 650, 651, 653, 69X : every subf exc . $6, $2, $= across these tags

No hyphens for X00, X10, and X11 fields 600, 610, 611, 700, 710, 711, etc), but hyphens for other fields.    
first non-empty value of 600$a, $d etc 610$a etc first non-empty value of 600$a,$d etc 610$a etc. , 610, 611, etc), but hyphens for other fields.    

Punctuation:
Follow punctuation and order given in
subfields:
600 - Subject Added Entry - Personal
Name (R)
610 - Subject Added Entry - Corporate Name (R)
611 - Subject Added Entry - Meeting
Name (R) , 610,611
(space between subfields) subfield subfield

Sample data:
600(subject - AuthorName) :
Kangxi,|cEmperor of China,|d1654-1722.
Facet/display:
kangxi,Emperor of China, 1654-1722.

Use double hyphen for subdivisions:
630 - Subject Added Entry - Uniform Title (R)
follow punctuation in fields, mainly using Periods between subfield value, except $v, $x, $y, $z
are "subdivisions" and use double - hyphen paired
with another subfield
Use double - hyphen for all subfields:
650 - Subject - Added Entry -  Topical Term (R)
651 - Subject Added Entry - Geographic  Name (R)
(double - hyphens between subfields) --
subfield -- subfield (period on end).
Sample data:
600(subject - AuthorName) :
Kangxi,|cEmperor of China,|d1654-1722.
Facet/display:
kangxi,Emperor of China, 1654-1722.
subfield

Sample data:
630(Standard title subject) :
Bible.|pN.T.|pLuke|xCommentaries.
Facet/display:
Bible.N.T.Luke -- Commentaries.

Use double - hyphen for all subfields:
 650, 651
(double - hyphens between subfields) --
subfield -- subfield (period on end).

Sample data:
650(subject-general):
Gardens|xSocial aspects|zChina|zBeijing|xHistory|y18th century.
Facet/display:
Gardens-Social aspectsChinaBeijingHistory-18th century.

All non-empty indexed
values.

Ordering of  multiple values:
Ordering of subject
headings in display
of individual bib
record should be in order of xml.

All non-empty  indexed values.

Hyphening pattern is same as display fields.

650(subject-general):
Gardens|xSocial aspects|zChina|
zBeijing|xHistory|
y18th century.
Facet/display:
Gardens--Socialaspects
-ChinaBeijing-
History--18th century.

Additional facets to be indexed:
1. Gardens - -Social aspects
2. Gardens - -Social aspects - -China
3. Gardens - -Social aspects - -China - -Beijing
4. Gardens - -Social aspects - -China - -Beijing - -History

Date of Publication

<marc:controlfield tag="008">[Date 1 in the 7-10 positions LR: Can also include 260 $c. (260-c is same as the value in control field. Use this if control field does not have pub date value.) (MV)does not have pub date value.) (MV).


Some records with 260 $c are having data other than date and data with no proper date.So for now 260 $c is ignored for publication date.

first value

all values

same as search field.

Keep decades, but add centuries and
"current year"
<controlfield tag="008">010108q10001999dcu b f000 0 eng d</controlfield>
The above example 1000-1999 means it should
appear with every facet 1000s,1100s, etc and every decade in 1000-1999.
Eliminate "odd" facets- don't want "s", "0000s",
or any years beyond the current year(manual removal 2020s and 9990s.
Only want single decades and single centuries.
<controlfield tag="008">001214k08000999xx zzz a n lat d</controlfield>.

Eg:
Two records with publication dates as 1905 and 1913.
The facets will be computed as:
20^th century (2)-- 1900 to 1999.
1900s (1)-- 1900 to 1909.
1910s (1) – 1910 to 1919.

Edition

250 - a,b (MV)

first value

all values

same as search field

Form/Genre

655 - a, v (MV)

first value

all values

same as search field

Language

<marc:controlfield tag="008">[language code in the 35-37 positions]</marc:controlfield> LR: Add 546 $a (MV)
Language Codes (iso-639-3)

all values

all values

same as search field

Format

856 - q
245 - h  

LR: Format is very tricky b/c many MARC fields/subfields can be used to determine format.  

I think we could also consider adding the following:

Leader 06/07 – these are Type of Record and Bibliographic Material – each is a single letter characters and when combined, they seem to map to the UPenn bib_format field.  (See the worksheet now named Format – marc)
007 Physical Description Fixed Field-General Information – the character positions indicate physical format information.  See [http://www.loc.gov/marc/bibliographic/bd007.html
] 655 $a, $v (Genre/Form)
300 $e, $3 (Extent: Accompanying Material, Materials specified)
337 $a (Media Type)
338 $a (Carrier Type)
340 $a, $e, $m, $3 (Physical Medium: Material base, support, book format, materials specified)
Possibly other 3XX fields/subfields.

In the end, we may want to just touch base with some SMEs (Gwyneth? Bob? Stuart) to determine specifically what we could/should include)
(MV)

first value

all values

same as search field

Manuscript

Has any holdings  with "manuscripts" in location_name (gets only this value)

LR: MARC XML does not have location_name so this is irrelevant to the IU data that  OLE has for November.  Manuscript could be determined by the Leader 06/07.  06 values a, f, t equal manuscripts on their own.  07 values c and d seem to imply mauscript/archival collections/series.  We should check with the SMEs on this one.

Microformat

Has 245 $h containing "micro" OR has any holdings  with "micro" in location_name OR call_number starts "micro" (gets only this value)

LR: the 245 $h "micro" will work for the IU OLE MARCXML we have, but the reamaing text is specific to UPenn.

Archive

Has any holdings  with "archive" in location_name (gets only this value)

LR: This is specific to UPenn.  We may need to talk to IU about if they include Archive descriptions in their MARC records and how they designate them as such.

Thesis/Dissertation

bib_format is 'tm' AND has a 502 field

LR UPenn's bib_format seems to be a combination of the data values found in the 06/07 Leader fields.  For example, t in the 06 is Manuscript and m in the 07 is Monograph/Item and together they equal a Thesis/Dissertation.

Conference/Event

Has a 111 or 711 field [LR: Include 611 or 811]

Book

bib_format is 'aa', 'am' or 'ac' or 'tm'; exclude $h [micro*] and $k [kit]

LR: the 2 characters are from the Leader 06/07 the inclusions are 245 subfields

Sound recording

bib_format is 'im' or 'jm' or 'jc' or 'jd' or 'js'

LR: the 2 characters are from the Leader 06/07

Musical score

bib_format is cm, dm, ca, cb, cd or cs

LR: the 2 characters are from the Leader 06/07

Map/Atlas

bib_format is 'e*' or 'fm'

LR: the 2 characters are from the Leader 06/07

Video

bib_format is 'gm' AND 007/0 = v

LR: the 2 characters are from the Leader 06/07

Projected graphic

bib_format is 'gm' AND 007/0 = g

LR: 007 is a controlled field that indicates the format/physical description at general level and then associated subfields are more specific.

Journal/Periodical

bib_format is 'as' or 'gs'

LR: 007 is a controlled field that indicates the format/physical description at general level and then associated subfields are more specific.

Image

bib_format is 'km'

LR: the 2 characters are from the Leader 06/07

Datafile

bib_format is 'mm'

LR: the 2 characters are from the Leader 06/07

Newspaper

bib_format is 'as' AND (008/21 = 'n' OR 008/22 = 'e' )

LR: the 2 characters are from the Leader 06/07.  The 008 controlled field in those 2 positions provides the "form")

3D object

bib_format is 'r*'

LR: the single character maps to the 06 position in the leader.

Database/Website

bib_format is '*i'

LR: the single character maps to the 06 position in the leader.

Government document

bib_format is NOT c*, d*, i*, j* AND ( (008/28 = f, i, o and 260$b not 'press') )

LR: the single character maps to the 06 position in the leader. 008 is a fixed length controlled field and 260 $b is a type of publication.

Other

any bib_format not caught above

LR: Presumably relates to other 06/07 Leader data values not represented.

Author

<dc:creator> 

<dcvalue element="contributor" qualifier="author">

first value

All non-empty
indexed values

All non-empty
indexed values

Description

<dc:description> (MV)
Per Bob P.: Show only <dc:description>.

Per Bob P.:  Do not show Abstract description.
[show blank]

first value

all values

same as search field

Language

<dc:language>  (MV)
Language Codes (iso-639-3)

<dcvalue element="language" qualifier="iso">en_US</dcvalue>
Language Codes (iso-639-1-cc)

first value

all values

same as search field

Subject

<dc:subject> (R)

<dcvalue element="subject" qualifier="none">

first value

all values

same as search field

Title

<dc:title>

<dcvalue element="title" qualifier="none">

first value

all values

same as search field

Type

<dc:type> (MV)

<dcvalue element="type" qualifier="none">

first value

all values

same as search field

Date of Publication

<dc:date>

<dcvalue element="date" qualifier="issued">

first value

all values

same as search field

Format

<dc:format> (MV)

<dcvalue element="type" (This is covered in a separate field. So do not include it in Format)
<dcvalue element="format" qualifier="mimetype">

??? (LR: In looking back at the MARC to Qualified DC mapping it is not entirely clear, but it should be both the  format and type elements.

first value

all values

same as search field

Publisher

<dc:publisher> (MV)

<dcvalue element="publisher"

??? (KG/LR: publisher.  It doesn't appear in the crosswalk, but that could be that the UMD dataset did not include that tag)

first value

all values

same as search field

ISBN/ISSN/other

<dc:identifier>(ISSN)0198-9669</dc:identifier>  (MV)
 <dc:identifier>(ISBN)0306710382</dc:identifier> (MV)

<dcvalue element="identifier" qualifier="isbn">0-918006-48-1</dcvalue>

first value

all values

same as search field