Skip to Main Content

A-Z Databases

Alphabets Find by Title

All A B C D E F G H I J K L M N O P Q R S T U V W X Y Z #
Video available
Audio available
Open Access
Accessibility info link
TexShare resource
GLAM resource

Standards and Practices for Creating Quality Digital Objects

These are the standards and practices followed by East Texas A&M University when creating a digital collection

Information about Metadata-Dublin Core


For more information about metadata, specifically Dublin Core, visit Dublin Core Metadata Initiative

Metadata


Metadata may be simply described as descriptive information about resources.  A metadata record contains prescribed elements that describe a record in some way that gives it value to the user of the collection.  There are several metadata standards currently in use, including MARC, METS, VRA Core, and Dublin Core.   OCLC’s CONTENTdm collection management system is the software we use to mount our collections on the Internet, and make the accessible to users.  CONTENTdm supports both the simple and qualified versions of Dublin Core.  A simple Dublin Core record is made up of 15 basic elements, whereas a qualified Dublin Core record contains 3 additional elements, plus several qualifiers for greater specificity.  Dublin Core was chosen due to its flexibility and relative simplicity.  It can easily describe a wide range of objects in various formats, and records can be easily created by non professional staff, if necessary.

Simple Dublin Core Elements:

Title

Format

Creator

Identifier

Subject

Source

Description

Language

Publisher

Relation

Contributor

Coverage

Date

Rights

Type

Qualified Dublin Core contains 3 additional fields: Audience, Provenance, and Rights Holder.  Each Dublin Core element is optional, and can be repeated if necessary. 

                It is important that digital item records have as much relevant information as possible to make them findable and useful.  Different collections may require different elements depending factors such as, the types of items in the collection, the audience, collection size etc.  A core set of elements that appear on all records establishes consistency.  The elements that must appear in all of our collections are:  Title, Subject, Description, Format, Date,  Publisher of Collection, and Rights.  Other fields may be added or repeated as necessary.

Formatting within metadata fields

Title:

The name by which a resource is formally known.  
Use the title given by the publisher or creator if available.  If not available, create a title.  
Written materials (prose, poetry) that don’t have official,  titles should use the first line of the item enclosed in brackets [I Don’t Like When You’re Sick]
Some author-created titles may be descriptive of the item (Letter from Bill to Steve)
If including dates in the title use ISO formatting.  
Follow capitalization rules from the MLA handbook unless the actual title of the material is specifically stylized.
Names within created titles should be in direct order (FirstName LastName)

Type

Format

Dates

ISO formatting in parentheses

Postcards and Images or other items with 2 sides

At the end of the title use (front) in parentheses for the front and (verso) in parentheses for the back.

Images with same descriptive title

Use a dash with a space before and after and a three-digit sequence number. Example: Cheerleaders in Formation – 001

Items that use a combination of the elements above

Use the elements above in descending order.

Example:

Photograph of Louise Tobin (1976-02-19) (front) – 001

Written materials that do not have an official title, date, author, or recipient noted.

First line enclosed in brackets.  Example:

[I Don’t Like When You’re Sick]

Capitalization from the MLA Handbook

The rules for capitalizing titles are strict. In a title or a subtitle, capitalize the first word, the last word, and all principal words, including those that follow hyphens in compound terms. Therefore, capitalize the following parts of speech:

  • Nouns (e.g., flowers, as in The Flowers of Europe)
  • Pronouns (e.g., our, as in Save Our Children; that, as in The Mouse That Roared)
  • Verbs (e.g., watches, as in America Watches Television; is, as in What Is Literature?)
  • Adjectives (e.g., ugly, as in The Ugly Duckling; that, as in Who Said That Phrase?)
  • Adverbs (e.g., slightly, as in Only Slightly Corrupt; down, as in Go Down, Moses)
  • Subordinating conjunctions (e.g., after, although, as if, as soon as, because, before, if, that, unless, until, when, where, while, as in One If by Land and Anywhere That Chance Leads)

Do not capitalize the following parts of speech when they fall in the middle of a title:

  • Articles (a, an, the, as in Under the Bamboo Tree)
  • Prepositions (e.g., against, as, between, in, of, to, as in The Merchant of Venice and “A Dialogue between the Soul and Body”)
  • Coordinating conjunctions (and, but, for, nor, or, so, yet, as in Romeo and Juliet)
  • The to in infinitives (as in How to Play Chess)

Creator:

The creator is the person, organization, or service who created the item.  Examples could include the author of a book, the person who took a photograph, or the person that painted a painting.
If creator is an individual, the name should be entered in reverse order (Last name, First name).  Do not include honorifics or titles such as Dr. or Mrs.

 

Subject:

The topic of the resource;

  • Subject describes what the resource is about, and generally provided the who, what, where, and when information of a resource 
  • Metadata creators may have to read the summary, abstract, table of contents, and/or skim the entire document to find appropriate subject headings and keywords
  • There are two types of subjects – subject headings and keywords
  • Subject headings come from a controlled vocabulary (i.e. LCSH) 
  • Keywords are added by the metadata creator and other contributors (resource creator, social tagging, etc.)
  • Associated Controlled vocabularies for subject headings:
  • Library of Congress Subject Headings (LCSH)
  • Library of Congress Name Authority File (LCNAF)
  • Subject headings must be written as found in LCSH or LCNAF
  • Use “--" between main heading and subheading

Example: Hunt County (Tex.)--Census, 1850

Publisher:

This field is only used to indicate that an item was published by a commercial entity.  Most often used for books, newspapers, and academic articles.
 

Description:

The description should be a natural language summary describing the object.  The description should have enough information to tell a researcher that this material is what they are looking for.

Description Type

About

Description of Physical Object

(All items)

A descriptive statement that includes specific characteristics about the object.

Examples: black & white photograph, handwritten letter, text document, color slide, black & white negative, color postcard, statue, lock of hair, plastic doll, an audio oral history, an oral history transcript, etc

Description of Condition

(All items with preservation concerns)

Condition of the physical materials

Examples: poor audio or video quality, damaged objects, objects with preservation concerns, brittle paper

Summary of Contents

(All items, but especially those with narrative content such as documents, audio oral histories, books, etc).

Rich, natural language summary describing the item. Include important details such as names of people and places, dates of events, major topics addressed, etc

Description for 2 sided objects such as photographs and postcards

Indicate on each record whether the user is looking at the front or verso.  The rest of the description should be the same as the other side.

Contributor:

The contributor field is used to give credit to the external entity, person, or institution that helped make the item available.  In many cases, the contributor field might refer to the donor of a particular collection. If contributor is an individual, the name should be entered in natural order (first name, last name)  If the contributor is unknown, leave the field blank
 

Date:

If date is known for an item or record, that information should be included.  There are several variations in the way that date information can be added to a record. Consistency is essential.  ISO 8601 should be followed whenever possible, when expressing dates. ISO 8601 compliant dates can be easily built and parsed by most programming languages. Examples of ISO compliant date forms include:

YYYY-MM-DD (2019-05-31)

YYYY (1977)

YYYY-MM (1977-06)

The ISO standard does not clearly state how to express inexact, or unknown dates. In cases where exact dates are not stated but it can be estimated, it should be expressed with circa, as in, circa YYYY, or circa YYYY-MM.. If there is no date listed, nor any reliable approximation, leave the field blank.

Type:

Still Image, Text, Collection, Dataset, Event, Image, Interactive Resource, Moving Image, Physical Object, Service, Software, Sound

File Format:

pdf, tif, cpd, f4v, jp2, jpg, mp3, mp4, mpg, URL, wav, xlsx

Identifier

In general, the identifier should refer to the name of the digital file, excluding the file extension.  This ensures that each item, especially those placed online, can be matched to a digital file (and, if it is a digitized item, a physical item and location).

Item

Needs Language Designation

Text based documents

Examples: news articles, books, handwritten letters, flyers, etc

Yes

Audio and video files with spoken or written language

Examples: oral histories, musical concerts that include a vocal element

Yes

Audio and video files without spoken or written language

Examples: silent football film, instrumental recordings

No

Postcards

Yes, if the back or front includes any typed or written language

Front of photographs (no text)

No

Verso of photographs (with written information, or a photographers stamp)

Yes

We use the MARC Code List for Languages to assign abbreviations to languages.  Our most common languages and their abbreviations are listed below:

Language

MARC Code

English

eng

Rights:

Rights are assigned using information from RightsStatements.org.  See Guide to Implementing Rights Statements from RightsStatements.org by the Society of American Archivists in order to assist with determining rights status.

Transcription:

Type

About

Words misspelled or other errors

use [sic] immediately after the error; may also include a tag or note in the record with corrected information

Words or portions of text that cannot be deciphered

[unintelligible] in place of the unknown text

Formatting

All documents are different so there may be some variation. Type what you see, and follow format and structure as closely as possible, but prioritize text over formatting.

Words that are crossed out

If text is crossed out, but legible use [crossed out] immediately after the text to note that it is marked through

Punctuation

Include punctuation exactly as it appears in the document

Additional Metadata Fields:

Interviewer:

Used for Oral History collections.  Names should be in reverse order (LastName, FirstName) and should not include honorifics.

Duration:

A/V collections will often contain a field for the duration of media.  Standard formatting is hour, minutes, seconds separated by colons.  Always include the leading zero(s) to express duration between 0 and 10 hours (00:58:38), (09:59:59)
Example: 00:36:29

 

Condition:

Although condition issues will usually be described in the notes field, some collections may contain enough condition concerns that an entire column is required.  Consult with the Digital Archivist.
 

Audio Quality:

When working with digitizing audio files, it may be necessary to include a column indicating the quality of the audio.  Although this information may not be presented in online digital collections, it is useful for making decisions about what files would go online.

  • Poor quality audio-Excessive background noise/hiss; dialog not distinguishable
  • Fair quality audio-Still may have significant background noise/hiss; more dialog can be understood
  • Good quality audio-May have a slight amount of background noise and hiss; dialog can be completely understood
  • Excellent quality audio-Very clean audio; No significant background noise or hiss


 

***It should be noted that this university hosts several collections for external institutions.  We provide server space and support for these collections, but all content decisions related to those collections including item selection, metadata etc. are made by each respective institution.