University of Minnesota
Immigration History Research Center

Immigration History Research Center's home page.


Development of COLLAGE

The project team has three primary goals:
  1. To facilitate undergraduate primary research in immigration studies;
  2. To provide public access to primary documents for K-12 education and for ethnic community organizations;
  3. To encourage the continued documentation and preservation of the immigrant experience.

As a resource, COLLAGE includes photographs, illustrations, text and other archival documents selected from over twenty-five different ethnic collections, with descriptive information for each image. The images will attract teachers, students, ethnic and community organizations, and the general public interested in immigration and ethnicity.

Support for COLLAGE comes from two grants. The first, from the University of Minnesota College of Liberal Arts Infotech Fees Committee, enabled the Center to begin the project. The second, a recent $250,000 award from the National Park Service’s heritage preservation program “Save America’s Treasures,” will allow it to expand this collection over the next several years. The expansive characteristics of the COLLAGE database provide that as more materials are selected, from the IHRC’s archives and other collections, and user interests are defined, the number of images available online through COLLAGE will increase.

The Digital Collection Unit of the University Libraries provides online support of the digital images. The IHRC’s COLLAGE is a partner in the campus-wide IMAGES initiative to produce high quality, standardized digital images and to promote their discovery.

For more information, please contact the IHRC's collection staff

News & Press Releases

For current information about COLLAGE, please look at our most recent IHRC News Online for updates and featured stories. The IHRC News Online is the monthly Web publication of the Immigration History Research Center.

Download the informational flyer on COLLAGE (PDF).

Current News

Finnish Radical Organizations' and Leaders' Photographs Available Online (Feb. 7, 2005)

Past COLLAGE Press Releases Include:

Articles Featuring COLLAGE:

  • "SR Visits: The Immigration History Research Center," by Carol Kelley in Serials Review vol. 29, no. 2 (2003): 151-153.
  • "COLLAGE," in Public History News, a quarterly publication of the National Council on Public History, vol. 23, no. 2 (Winter 2003): 13.
  • "A 'COLLAGE' of Digitized Primary Sources Documenting Immigration is Accessible Online," in Information Technology Newsletter, published by the Office of Information Technology, University of Minnesota, vol.7 (Jan. 2003): 4.
  • "News from the Midwest - Immigration History Research Center, University of Minnesota," in MAC Newsletter, a publication of the Midwest Archives Conference, (Jan. 2003): 20-21.
  • "IHRC Offers Searchable Online Image Database," in Twin Cities Archives Round Table Newsletter vol. 18, no. 1 (June 2002): 5-6.

Return to Top

Metadata Standards and Technical Information

The following information details the digitization process, metadata creation and XML conversion for items selected to be digitized.

Digitization Process

All image files are scanned from the original object or high quality facsimile held within the collection. The quality of each scan is defaulted to 300 ppi (pixels per inch) resolution, 24-bit RGB color, on a one to one (100%) ratio. The creation hardware is an Epson Expression 836 XL scanner and internal Epson TWAIN Pro version 2.10A driver on a Windows enabled Dell Dimension 4550 desktop running Microsoft XP Professional. The creation software is Adobe Photoshop CS (version 8.0). The files are saved as uncompressed, PC byte order tiff images and identified using a unique identifier. All tiff images are stored offline on CD-R.

Derivative image files for Web display are created using Adobe Photoshop CS. Web reference images are resized from the original tiff image to 600 pixels in width and 96 ppi resolution. The reference files are saved as a medium-quality jpeg image using the same unique identifier as the corresponding tiff image and stored online in a designated folder. Thumbnail images are resized from the original tiff image to 140 pixels in width and 72 ppi resolution. The thumbnail files are saved as a low-quality jpeg image using the same unique identifier as the corresponding tiff image and stored online in a designated folder separate from the reference images.


As individual items are scanned, descriptive and technical information is cataloged in a Microsoft Access 2003 (XP) database. The database tracks three types of metadata for each digital asset:

1. Descriptive metadata: unique information about the digital object used for indexing and identification
2. Structural metadata: relational aspects of the digital object to the collection
3. Administrative metadata: technical data on the image file, rights management, and preservation

The individual metadata elements follow the standards outlined in the Western States Dublin Core Metadata Best Practices, version 2.0 (January 2005). Most fields map to one of the 15 metadata elements established as the Dublin Core Metadata Element Set version 1.1 (ANSI/NISO Z39.85-2001).

Controlled vocabularies provide consistency in data entry and record searching. Subject headings follow either the Library of Congress Subject Headings (LCSH) authority terms or adhere to LCSH format. Naming and descriptive conventions follow recommended practices in the Anglo-American Cataloging Rules 2nd edition (AACR2). Language encoding uses the MARC three letter code, not the international standard ISO 639 recommended by the Western States Dublin Core.

Crosswalk map of IHRC descriptive metadata fields and applicable schemes, Western States labels and Dublin Core:

IHRC Descriptive SchemaWestern States Dublin CoreDublin Core
Digitalid (scheme="local") Identifier DC.Identifier
Standardid (scheme="local") Identifier DC.Identifier
Title Title DC.Title
Caption Description DC.Description
Annotation Description DC.Description
Creator (scheme="AACR2") Creator DC.Creator
Contributor (scheme="AACR2") Contributor DC.Contributor
Imageyear Date.Original DC.Date
Imageperiod Date.Original DC.Date
Medium Type DC.Type
Dimensions Description DC.Description
Language (scheme="MARC") Language DC.Language
Description Description DC.Description
Subject (scheme="LCSH," "local") Subject DC.Subject
Note Description DC.Description

Crosswalk map of IHRC structural metadata fields and applicable schemes, Western States labels and Dublin Core:

IHRC Structural Schema Western States Dublin Core Dublin Core
Series Relation.IsPartOf DC.Relation
Holding (scheme="AACR2") Contributing.Institution n/a
Provenance Relation.IsPartOf DC.Relation
Object.Location n/a n/a
Published Source DC.Source
Relationtitle Relation.IsPartOf DC.Relation
Relatedurx (scheme="URI") Relation.IsPartOf DC.Relation

Crosswalk map of IHRC structural metadata fields and applicable schemes, Western States labels and Dublin Core:

IHRC Structural Schema Western States Dublin Core Dublin Core
Format (scheme="MIMIE") Format DC.Format
Filesize (in bytes) Digitization.Specifications.Filesize DC.Description
Resolution (ppi) Digitization.Specifications.Quality DC.Description
Bitdepth Digitization.Specifications.Quality DC.Description
Extent (in pixels h x w) Digitization.Specifications.Extent DC.Description
File.Created Date.Digital.Created DC.Date
File.Modified Date.Digital.Modified DC.Date
File.Exported Date.Digital.Issued DC.Date
Userights Rights DC.Rights
Rights.Management Rights DC.Rights

XML Conversion

The Access database is regularly converted into an XML file using XMetaL version 3.0 software. A query in the Access database creates a new table consisting of the fields required for conversion. An internal macro in XMetaL connects to this queried Access table via the ODBC data source driver in Windows XP. XMetaL extracts the data from the database fields and validates the elements against the DTD (Document Type Definition) file written specifically for the IMAGES project. Once validated, the Access records are successfully translated into XML markup. The XML file is uploaded onto a server where it is indexed for searching and linked to the Web reference image files for the corresponding record. This is then searchable via a search portal on our website.