Inside CDL
Inside CDL | CDL Home | Search | Contact the CDL
Collection Development Process
Collections
Projects, Programs & Services
Committees & Groups
Digital Library Building Blocks
Instructional Materials
Assessment
News & Events
Report a Problem
Contact the CDL
Digital Library Building Blocks > CDL Guidelines for Digital Objects

Appendix A. Descriptive Metadata Guidelines (Detailed)

Table of Contents

The following conventions are used to express guidelines for each metadata element:

  • Definition: A definition of the element.
  • Recommended data values: Recommended data values for the element. May include references to appropriate content standard, authority file, thesaurus, encoding standard, etc. to guide data value entry.
  • Crosswalks: The crosswalks provide encoding analogs between elements in Dublin Core and MODS, two schemas for descriptive metadata that are commonly used with METS.
  • Examples: Provides examples of preferred data values within elements.

Identifier

Definition:A unique identifier for the resource.

Recommended data values: Identify the resource by means of a unique string or number conforming to a formal or locally-derived identification system. Example formal identification systems include:

  • Uniform Resource Identifier (URI), including the Uniform Resource Locator (URL)
  • Digital Object Identifier (DOI)
  • International Standard Book Number (ISBN)

Crosswalks:

  • Dublin Core:
    • <dc:identifier>
  • MODS:
    • <identifier type="">
    • <location>
           <url>

Examples:

calb_p3353 [Note: locally-derived unique identifier]
0609609718 [Note: ISBN]

Title

Definition: A succinct, identifying name for the resource.

Recommended data values: Transcribe the formal title of the resource or supply a title, if necessary, using an appropriate content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), Describing Archives: a Content Standard (DACS), or Graphic Materials (GIHC).

Crosswalks:

  • Dublin Core:
    • <dc:title>
  • MODS:
    • <titleInfo>
           <title>
    • <titleInfo>
           <title>
                <subtitle>

Examples:

Formal titles
Two dancers on a stage / Frasher Foto [Note: transcribed according to AACR2]
The Rocky Mountains, emigrants crossing the plains [graphic] / F.F. Palmer, del. [Note: transcribed according to Graphic Materials]

Supplied titles
[Photograph of musicians performing at a cultural program] [Note: derived according to AACR2]
Mitchell Bonner photograph of musicians performing at a cultural program [Note: derived according to DACS]
[Phoenix] / Ben Shahn [Note: derived according to Graphic Materials]

Creator

Definition: The name of the person, institution, agent, or group primarily responsible for the creation of the resource.

Recommended data values: The form of the name should be taken from a standard naming authority file, such as the Library of Congress Name Authority File (LCNAF) or Union List of Artists' Names (ULAN). If a name does not appear in an authority file, establish the name according to a content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), or Describing Archives: a Content Standard (DACS).

Additionally, if possible, indicate the code for a standard naming authority file from which the name is taken. Use "lcnaf" for the LCNAF or "ulan" for ULAN. For all others, use the appropriate code for the source (see the Library of Congress' Term, Name, and Title Sources Code List).

If the name is not found in a standard naming authority file, indicate the content standard by which the name is established, e.g., "aacr" for AACR2, "dacs" for DACS, and "gihc" for Graphic Materials (GIHC) (see the Library of Congress' Descriptive Conventions Code List). If a content standard is not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:creator>
  • MODS:
    • <name type="personal | corporate | conference" authority="">
           <namePart>
    • <name type="personal | corporate | conference" authority="">
           <namePart>
                <role>
                     <roleTerm type="text">

Examples:

Personal name entry
Yamada, Mitsuye [Note: determined from local cataloging authority or LCNAF]
Chase, Alexander W. (Alexander Wells), 1843-1888 [Note: derived according to AACR2]
White, Ira Johnson [Note: determined from ULAN]
Robinson family [Note: derived according to DACS]

Corporate name entry
American Philosophical Society [Note: determined from local cataloging authority or LCNAF]
Frasher Foto (Firm) [Note: derived according to AACR2]

Date

Definition: A single date or inclusive dates indicating when the resource was created.

Recommended data values: Construct dates using an appropriate content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), Describing Archives: a Content Standard (DACS), or Graphic Materials (GIHC).

At least one form of the date must be normalized (note that the Date element is repeatable) using one of the following standards:

Crosswalks:
  • Dublin Core:
    • <dc:date>
    • <dcterms:created>
  • MODS:
    • <originInfo>
           <dateCreated encoding="temper | w3cdtf" qualifier=""> [Note: do not use <dateCaptured> when describing date of creation for a born-digital resource]
    • <originInfo>
           <dateOther encoding="temper | w3cdtf" qualifier="">
    • <publicationInfo>
           <dateIssued encoding="temper | w3cdtf" qualifier="">

Examples:

TEMPER encoding

Single dates

1901 = 1901
January 1901 = 19010100
1901 January 3 = 19010103

Date spans

1900-1950 = 1900-1950
1956 January-July = 19560100-19560700
1980s = 1980-1989 [Note: use an interval to indicate every year of the decade]
19th century = 1801-1900

Broken date spans

1924, 1956-1975 = 1924, 1956-1975 [Note: separate by a comma]

Open date spans
1911- = 1911-
-1911 = -1911

Approximate dates
circa 1950 = 1950~

Undated material

undated: circa mid 20th century = 1935~-1965~ [Note: if a resource is undated this can be stated but provide an estimate if possible; normalize as an interval, perhaps using the dates of the life of creator, etc.]

International Standard Organization (ISO) 8601 encoding (using a modified version of the W3C date and time formats profile)

Single dates

1901 = 1901
January 1901 = 1901-01
1901 January 3 = 1901-01-03

Date spans

1900-1950 = 1900/1950
1956 January-July = 1956-01/1956-07
1980s = 1980/1989 [Note: use an interval to indicate every year of the decade]
19th century = 1801/1900

Broken date spans

1924, 1956-1975 = 1924, 1956/1975 [Note: separate by a comma]

Open date spans
1911- = 1911/9999 [Note: use an interval and set the end date to 9999]

Approximate dates
circa 1950 = 1945/1955 [Note: normalize as an interval to express an appropriate date range]

Undated material

undated: circa mid 20th century = 1935/1965 [Note: if a resource is undated this can be stated but provide an estimate if possible; normalize as an interval, perhaps using the dates of the life of creator, etc.]

Description

Definition: A brief free-text note, abstract, table of contents listing, or descriptive statement that characterizes more fully than the title does the scope or content of the resource.

Recommended data values: Use when the intellectual content of the item is not sufficiently captured in the title and other descriptors. Construct a note using an appropriate content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), Describing Archives: a Content Standard (DACS), or Graphic Materials (GIHC).

Crosswalks:
  • Dublin Core:
    • <dc:description>
    • <dcterms:abstract>
  • MODS:
    • <abstract>
    • <tableOfContents>
    • <note>
    • <note type=""> [Note: use for scope and content notes that are equivalent to MARC 520 element]

Examples:

Depicts unknown automobile driver stopping at roadside to add water to engine on all-day drive from Chico to Sacramento. Exact location unknown. Verso stamped with 596; manuscript note indicates car owned by "N.E.R." [Note: derived according to AACR2]
View of the Alaskan King Ice Cream Parlor, with horse-drawn delivery wagon in foreground and City Hall in background, Eugene, OR. [Note: derived according to DACS]
Signed in red ink. Edition of 59; Library has 14/59. [Note: derived according to Graphic Materials]

Language

Definition: Term that indicates the language that is an integral part of the resource, such as a caption that is part of a photograph or a title that is part of a painting.

Recommended data values: At least one form of the language term must be normalized in coded form using a three-letter code from the from the International Organization for Standardization (ISO) 639-2 Codes for the Representation of Names of Languages (note that the Language element is repeatable, for representing the language term in textual form).

Crosswalks:
  • Dublin Core:
    • <dc:language>
  • MODS:
    • <languageTerm authority="iso639-2b" type="code">

Examples:

eng [Note: use for English]
vie [Note: use for Vietnamese]
ger [Note: use for German]

Subject (Name)

Definition: Significant names (personal, corporate, family, meeting, etc.) represented in or by the resource.

Recommended data values: The form of the name should be taken from a standard naming authority file, such as the Library of Congress Name Authority File (LCNAF) or Union List of Artists' Names (ULAN). If a name does not appear in an authority file, establish the name according to a content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), or Describing Archives: a Content Standard (DACS).

Additionally, if possible, indicate the code for a standard naming authority file from which the name is taken. Use "lcnaf" for the LCNAF or "ulan" for ULAN. For all others, use the appropriate code for the source (see the Library of Congress' Term, Name, and Title Sources Code List).

If the name is not found in a standard naming authority file, indicate the content standard by which the name is established, e.g., "aacr" for AACR2, "dacs" for DACS, and "gihc" for Graphic Materials (GIHC) (see the Library of Congress' Descriptive Conventions Code List). If a content standard is not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:subject>
  • MODS:
    • <subject authority="">
           <name type="personal | corporate | conference" authority="">
                <namePart>

Examples:

Personal name entry
Yamada, Mitsuye [Note: determined from local cataloging authority or LCNAF]
Chase, Alexander W. (Alexander Wells), 1843-1888 [Note: derived according to AACR2]
White, Ira Johnson [Note: determined from ULAN]
Robinson family [Note: derived according to DACS]

Corporate name entry
American Philosophical Society [Note: determined from local cataloging authority or LCNAF]
Frasher Foto (Firm) [Note: derived according to AACR2]

Subject (Title)

Definition: Significant titles of other resources (e.g., works, expressions of those works, individual items, etc.) represented in or by the resource.

Recommended data values: The form of the title should be taken from a standard naming authority file, such as the Library of Congress Title Authority File (LCTAF). If a title does not appear in an authority file, establish the title according to a content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), Describing Archives: a Content Standard (DACS), or Graphic Materials (GIHC).

Additionally, if possible, indicate the code for a standard naming authority file from which the title is taken. Use "lctah" when the name is established in the LCTAF.

If the title does not appear in the authority file, indicate the content standard by which the title is established, e.g., "aacr" for AACR2, "dacs" for DACS, and "gihc" for Graphic Materials (GIHC) (see the Library of Congress' Descriptive Conventions Code List). If a content standard is not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:subject>
  • MODS:
    • <subject authority="">
           <titleInfo authority="">
                <title>

Examples:

Kim Hà, 1950-. Qua con bao du : hoi ky vuot bien bang duong bo. [Note: manuscript material documenting creation of a monograph titled "Qua con bao du"; entry derived according to AACR2]

Subject (Place)

Definition: Significant names of geographic locations represented in or by the resource.

Recommended data values: The form of the name should be taken from a standard naming authority file, such as the Library of Congress Name Authority File (LCNAF) or Thesaurus of Geographic Names (TGN). If a name does not appear in an authority file, establish the name according to a content standard such as Anglo-American Cataloging Rules (AACR2) or Cataloging Cultural Objects (CCO).

Additionally, if possible, indicate the code for a standard naming authority file from which the name is taken. Use "lcnaf" for the LCNAF or "ulan" for ULAN. For all others, use the appropriate code for the source (see the Library of Congress' Term, Name, and Title Sources Code List).

If the name is not found in a standard naming authority file, indicate the content standard by which the name is established, e.g., "aacr" for AACR2, "dacs" for DACS, and "gihc" for Graphic Materials (GIHC) (see the Library of Congress' Descriptive Conventions Code List). If a content standard is not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:coverage>
    • <dcterms:spatial>
  • MODS:
    • <subject authority="">
           <geographic>
    • <subject authority="">
           <hierarchicalGeographic>
    • <subject authority="">
           <cartographics>

Examples:

Santa Cruz (Calif.) [Note: determined from local cataloging authority or LCNAF]
Santa Cruz [Note: determined from TGN]
Rancho Boca de la Playa (Calif.) [Note: established according to AACR2]

Subject (Topic, Function, or Occupation)

Definition: Significant topics or subjects (including concepts, events, etc.), functions, or occupations represented in or by the resource.

Recommended data values: The form of the heading should be taken from a standard or local thesaurus, such as the Library of Congress Subject Headings (LCSH), Art and Architecture Thesaurus (AAT), or Thesaurus of Graphic Materials I (TGM I).

If a heading does not appear in a thesaurus, establish the heading according to standard thesaurus rules (such as the Library of Congress' Subject Cataloging Manual, AAT rules, or TGM I rules), or local thesaurus rules.

Additionally, if possible, indicate the code for a standard naming authority file from which the heading is taken. Use "lcsh" for LCSH, "aat" for AAT, or "gmgpc" for TGM I (see the Library of Congress' Term, Name, and Title Sources Code List).

If the heading does not appear in a standard thesaurus, indicate the thesaurus rules by which the term is established, e.g., "lcsh" for LCSH, "aat" for AAT, or "gmgpc" for TGM I (see the Library of Congress' Term, Name, and Title Sources Code List). If standard thesaurus rules are not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:subject>
  • MODS:
    • <subject authority="">
           <topic>
    • <subject authority="">
           <occupation>

Examples:

Viticulture -- California -- Sonoma County [Note: determined from LCSH]
Surveyors--California--Orange County [Note: determined from LCSH]
Street railroads [Note: determined from AAT]
Agricultural laborers--Italian Americans--California--Salinas [Note: determined from TGM I]

Genre

Definition: Primary genre(s) represented in or by the resource.

Recommended data values: The form of the heading should be taken from a standard or local thesaurus, such as the Library of Congress Subject Headings (LCSH), Art and Architecture Thesaurus (AAT), Genre Terms (RBGENR), or Thesaurus of Graphic Materials II (TGM II).

If a heading does not appear in a thesaurus, establish the heading according to standard thesaurus rules (such as the Library of Congress' Subject Cataloging Manual, AAT rules, or TGM II rules), or local thesaurus rules.

Additionally, if possible, indicate the code for a standard naming authority file from which the heading is taken. Use "lcsh" for LCSH, "aat" for AAT, or "gmgpc" for TGM II (see the Library of Congress' Term, Name, and Title Sources Code List).

If the heading does not appear in a standard thesaurus, indicate the thesaurus rules by which the term is established, e.g., "lcsh", "aat", or "gmgpc". Use the appropriate code for the thesaurus by which the term is established (see the Library of Congress' Term, Name, and Title Sources Code List). If standard thesaurus rules are not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:type>
  • MODS:
    • <genre authority="">

Examples:

Photographs [Note: determined from LCSH]
Photographic prints [Note: determined from AAT]
Photographic prints [Note: determined from TGM II]

Type

Definition: A high-level type data value that generally characterizes the resource represented by the digital object. This high-level data value may also be repeated, or more specific genre data values may also be encoded as part of the descriptive metadata (see Genre).

Recommended data values: Choose data values from one of the following lists, based on the descriptive metadata scheme being utilized. Select data values from the MODS type vocabulary if in doubt:

Crosswalks:
  • Dublin Core:
    • <dc:type>
  • MODS:
    • <typeOfResource>

Examples:

image [Note: determined from Dublin Core type vocabulary]
still image [Note: determined MODS type vocabulary]

Format/Physical Description

Definition: The physical or digital manifestation of the resource. Typically, this may include the media-type or dimensions of the resource. Examples of dimensions include size and duration.

Recommended data values: Construct a statement using an appropriate content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), Describing Archives: a Content Standard (DACS), or Graphic Materials (GIHC).

Crosswalks:
  • Dublin Core:
    • <dc:format>
    • <dcterms:extent>
  • MODS:
    • <physicalDescription>
           <extent>
    • <physicalDescription>
           <form>
    • <physicalDescription>
           <internetMediaType>

Examples:

1 photographic print ; 9 x 14 cm. [Note: derived according to AACR2]
14 letters [Note: derived according to DACS]
1 leaflet : ill. ; 21.5 x 38.5 cm., folded to 21.5 x 10 cm. [Note: derived according to Graphic Materials]

Related Collection/Project

Definition: A machine access-oriented identifier for a collection or project that the resource is a member of or related to in some manner.

Recommended data values: If the resource is a member of or related to a collection or project, at least one Related Collection/Project element must refer to a unique identifier or to the title for the collection or project. Data producers may use a unique local identifier to refer to a related collection or project, such as a URL or other unique identifier. Unique identifiers managed by the CDL are preferred. To obtain a unique CDL identifier for a particular collection or project, contact the CDL.

Alternatively, indicate the title for a collection or project. Note that use of free-text data such as this will not guarantee that the resource can be uniquely subset by CDL-hosted collections that it is a part of.

Crosswalks:
  • Dublin Core:
    • <dc:relation>
    • <dcterms:isPartOf>
  • MODS:
    • <relatedItem>
           <url>
    • <relatedItem>
           <identifier>

Examples:

CDL-supplied unique identifier
http://jarda.cdlib.org

Title
Silicon Valley History Online

Institution/Repository

Definition: The name of the owning or contributing institution of the resource.

Recommended data values: The form of the name should be taken from a standard naming authority file, such as the Library of Congress Name Authority File (LCNAF). If the name does not appear in an authority file, establish the name according to a content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), or Describing Archives: a Content Standard (DACS).

Crosswalks:
  • Dublin Core:
    • <dc:publisher>
  • MODS:
    • <location>
           <physicalLocation>

Examples:

Fowler Museum of Cultural History [Note: determined from local cataloging authority or LCNAF]
Orange Public Library [Note: derived according to AACR2]
University of California, San Francisco. Library. Archives and Special Collections [Note: derived according to AACR2]

Contributor

Definition: The name of the person, institution, agent, or group responsible for contributing to the resource in some significant manner, such as a illustrator, designer, autographer, etc.

Recommended data values: The form of the name should be taken from a standard naming authority file, such as the Library of Congress Name Authority File (LCNAF) or Union List of Artists' Names (ULAN). If a name does not appear in an authority file, establish the name according to a content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), or Describing Archives: a Content Standard (DACS).

Additionally, if possible, indicate the code for a standard naming authority file from which the name is taken. Use "lcnaf" for the LCNAF or "ulan" for ULAN. For all others, use the appropriate code for the source (see the Library of Congress' Term, Name, and Title Sources Code List). .

If the name is not found in a standard naming authority file, indicate the content standard by which the name is established, e.g., "aacr" for AACR2, "dacs" for DACS, and "gihc" for Graphic Materials (GIHC) (see the Library of Congress' Descriptive Conventions Code List). If a content standard is not used, use "local".

Crosswalks:
  • Dublin Core:
    • <dc:contributor>
  • MODS:
    • <name type="personal | corporate | conference" authority="">
           <namePart>
                <role>
                     <roleTerm type="text">

Examples:

Personal name entry
Yamada, Mitsuye [Note: determined from local cataloging authority or LCNAF]
Chase, Alexander W. (Alexander Wells), 1843-1888 [Note: derived according to AACR2]
White, Ira Johnson [Note: determined from ULAN]
Robinson family [Note: derived according to DACS]

Corporate name entry
American Philosophical Society [Note: determined from local cataloging authority or LCNAF]
Frasher Foto (Firm) [Note: derived according to AACR2]

Publisher

Definition: The name of the publisher of a formally published resource. This element may not be relevant for unpublished materials.

Recommended data values: The form of the name should be taken from a standard naming authority file, such as the Library of Congress Name Authority File (LCNAF) or Union List of Artists' Names (ULAN). If a name does not appear in an authority file, establish the name according to a content standard such as Anglo-American Cataloging Rules (AACR2), Cataloging Cultural Objects (CCO), or Describing Archives: a Content Standard (DACS).

Crosswalks:
  • Dublin Core:
    • <dc:publisher>
  • MODS:
    • <originInfo>
           <publisher>

Examples:

Simon & Schuster [Note: determined from local cataloging authority or LCNAF]
New Albion Records [Note: derived according to AACR2]

Document Owner: California Digital Library

Last reviewed: September 26 , 2007

URL: http:// www.cdlib.org/inside/diglib/guidelines/dmdreqs.html

Contact the CDL
Questions? Comments?

Copyright © 2008 The Regents of the University of California