IMLS Proposal:
Connecticut History Online Phase Two

June, 2002

Table of Contents

Abstract

Narrative

• National Impact
• Adaptability
• Design
• Management Plan
• Budget
• Personnel
• Evaluation
• Dissemination
• Contributions
• Sustainability

1. The materials can be searched by key word and advanced keyword searches by subject, creator, date, place name, collection, institution, and title.
2. Materials can also be searched using the unique GeoLocator that allows the user to view a map of Connecticut and seek materials from a particular area or city. The feature allows the user to continually narrow or widen the search of the map getting into finite detail of towns and other geographic features.
3. Materials can also be found through the use of defined searches entitled “Journeys” that allow the user to review material related to a pre-defined topic that places the material into an historical context.
4. There will be a fourth link to material via collection-level cataloging records that will be contributed to the OCLC database, widening Connecticut History Online’s accessibility.

The project will continue to build resources that can be used by secondary school teachers and students, but will also include scholarly researchers, diverse Web users, and other museum and library professionals. As the project expands, we hope that it will provide a model to enhance access to historical digital collections for other statewide coalitions.

Back to top

• Continue and expand upon a model of inter-institutional cooperation that will serve as an example for other groups.
• Create a large virtual collection of research materials images from five institutions distant from one another.
• Expand the consortial metadata catalog to include manuscripts, broadsides, audio clips, oral history transcripts, artifacts, maps, diaries and publications as well as graphics.
• Utilize standard technologies to disseminate and share the data captured with the national community of scholars, researchers, and the general public in a comprehensible manner.
• Promote the use of collections in electronic formats using national standards in linked databases.
• Continue and expand upon the innovative GeoLocator that links collections to an international geographic information system (GIS).
• Encourage broader use of AAT, the Getty Information Institute’s Art and Architecture Thesaurus by promoting the educational use of digital images.
• This aspect of the project will also encourage best practices in the museum and library field given that it will: Involve representatives of the target user community, middle and high school teachers and students, as project co-researchers/investigators.
• Encourage inquiry-based learning by promoting the use of original source material.

Connecticut History Online will offer the obvious benefit of joining the resources of five institutions. Since it will not be site specific, it will be available to the widest possible audience anywhere there is access to the Web, 24 hours per day, seven days per week. By assembling a large mass of 19th - and early 20th -century materials, Connecticut History Online will provide research opportunities not available anywhere else – not in any single repository and not even easily at all five of the collection-holding partners bringing together multiple formats from a variety of institutions allows the researcher to draw from library, museum and historical society collections. The convenience of using this single collection at any time of the day or night from anywhere in the world means that Connecticut History Online will receive much more use than the parts of the collection now receive at the participating institutions.

Connecticut History Online has, and will continue, to serve as a national model for cooperation among museums and libraries, the use of electronic media to deliver historical materials, and as a medium that supports both secondary students and life-long learners. In a geographic region not known for cooperation, CHO has overcome New England's reputation for rugged independence that resulted in a truly cooperative effort.

Phase Two of CHO will build upon a cooperative model of shared responsibility, equal project ownership, clear and consistent communication, and the individual strengths of each of the partners. The five institutions will jointly own CHO. However, each partner retains individual ownership of its images and whatever copyright will reside in those images. To successfully plan and carry out the project, the partners will use management, cataloging, and website planning teams made up of representatives of each of the five participating institutions. Building on past success, communication will be by listserve and regular meetings.

The partnership of museums and libraries will continue to develop a strong synergy. Phase One of the project focused on one type of material common to all institutions: graphics. Phase two will significantly expand that material to reflect the diversity of the partners and items include manuscripts, broadsides, audio clips, oral history transcripts, artifacts, maps, diaries, maps, artifacts, and costumes.

Phase Two will build upon the common strength of the partners. The project will continue to use library-based technology and metadata that grows out of a library tradition, and the site will be mounted on servers at the University of Connecticut, a major research library. At the same time, the project will draw upon the experience and expertise of Mystic Seaport and The Connecticut Historical Society, a museum and a historical society who regularly work with teachers and students and in developing curriculum materials. Staff from each of the five institutions brings specialized knowledge and subject expertise that allows them to manage and describe these specialized holdings.

Back to top

ADAPTABILITY

• Significance of the subject matter in Connecticut and American history, the relevance of the images to contemporary research interests and teaching strategies as well as the match of the materials to the subject areas listed above.
• Geographic coverage of different parts of Connecticut.
• Uniqueness of the material; although it is assumed that there will be little overlap in the five institutions’ collections.
• Copyright has been cleared. Material will only be used if copyright has expired, the institution has obtained copyright release, or the institution owns copyright.

The Connecticut Historical Society proposes adding material from four new genres: broadsides, maps, costumes, and oral histories. It will include 1,500 of its broadsides from a collection that numbers more than 3,500 items. It will focus on the period 1776 to 1900 and covering such subjects as music, the military, Native Americans, and African Americans. It will also add 500 Connecticut maps with the major focus on the era from 1826 to 1925 of which approximately 75% are unique and not found in Edmund Thompson’s Maps of Connecticut. It will add 250 images of historical costumes out of a total of 5,300 items covering the period 1760 to 1890 with a major focus on middle and working class apparel. Finally, CHS will add ten oral interviews of West Indians living in Hartford during the period of 1940 to 2000. There materials will be presented with a one-minute sound clip from the tape along with a copy of the transcript and a summary.

Mystic Seaport will focus on digitizing manuscript materials, nautical charges, oral histories and museum artifacts. It will draw on its extensive collection of ships logs and journals and personal diaries from the 19th century for approximately 1,200 of the items it will select. These materials reflect such activities as fishing, whaling and trading done by Connecticut ships and companies. Mystic will also digitize 30 nautical charts from coastal Connecticut and include 35 oral histories done in 1993-1994 with families connected with the fishing industry in Connecticut. Finally, Mystic will digitize more than two hundred artifacts from its renowned collection of more than 20,000 paintings, scrimshaw, carvings, model vessels, tools and other objects.

The Thomas J. Dodd Center will add a total of 1,500 items from its collection. Material will come from its Alternative Press Collection (200 items) and focus on the anti-war movement and African American and Women’s activist groups of the 1960s and 1970s in Connecticut. It will add items from the Southern New England Telephone Company archives, one of America’s first telephone companies and creator of the world’s first telephone exchange, (550 items) that includes broadsides, early telephone directories including the first directory ever produced in the world, equipment diagrams, artwork, artifacts, and town histories created by the company. Other material will be drawn from the Dodd Center’s railroad collection (400 items) and include railroad and trolley line maps, railroad station blueprints, board of director’s minutes of 19th century railroad companies. Finally, the Dodd Center will draw on materials from its Connecticut History and University Archives materials (250 items) focusing on Connecticut during the war years (1917-1972), Connecticut’s congressional delegation, and University athletics.

The Connecticut State Library will add 2,500 items to CHO. It will digitize the 500 photographic images from the Lewis Sprague Mills Collection. The Mills Collection documents rural Connecticut schools as well as bridges, historic sites, public buildings, and mills that were constructed during the 17th and 18th centuries. The State Library will also add 2,000 items from the military services questionnaires completed by Connecticut men and women following World War I. These documents were gathered to document Connecticut’s wartime contributions and include personal data, war service and contain images of each of the individuals. The materials have significant research value for the study of ethnicity and for genealogy.

The New Haven Colony Historical Society will add 1,500 New Haven Redevelopment Agency photographic images to CHO from its collection of more than 12,000 items. These cover the period 1953 to the early 1980s. The images include streetscapes and images of individual buildings often showing before and after images of specific locations. The material documents an era of change that occurred in Connecticut cities and cities across the country as the federal government undertook major programs to “renew” the urban environment.

Use of National Standards

This project will continue to adhere to the national standards established in Phase One of the project for data capture, description, and display. Catalogers will used the manual created during Phase One which is based on MARC standards for metadata files. The project will use nationally recognized thesauri for assignment of subject headings within the descriptive records. The Art and Architecture Thesaurus (AAT) and the Library of Congress Subject Headings (LCSH) will be the primary thesauri used. The forms of personal and corporate names used will be verified in the Name Authority File maintained by the Library of Congress. The project will also continue to use the specialized digital gazetteer that was especially created for this project during Phase One.

The project will use Endeavor Voyager, an integrated library system, to manage the database and metadata records. Endeavor Voyager has become an industry standard for ILS systems at most major research libraries throughout the United States. CHO will continue operate as a separate database located on a University of Connecticut server. Endeavor is fully Z39.50 compliant, allowing multiple site searching. The project will also continue to use Endeavor ImageServer software. ImageServer maintains information about the location of captured data and is capable of delivering images, documents, maps, audio, and video.

The Geolocator is a unique feature of CHO Phase One. In an attempt to make searching a bibliographic database more appealing to a wide audience, project staff wanted to create a map of Connecticut with dynamic links between images and their records. This required evaluating the efficacy and structure of the geographic fields in a MARC record for use with a Geographic Information System. Using an ambiguous search (searching for all instances of a name), the Geographic Name field (651) was inefficient. By pre-processing all the names in the 651 field, a more specific search was achieved. Tables were created after an initial polling of the geographic names in the database. URLs are constructed from the place names and then geo-referenced against appropriate thesauri using ArcView. To take into account historical names and places that no longer exist, the Connecticut Digital Gazetteer was created as one of the GeoLocator’s thesauri. The results of searches produce a Connecticut map that includes graphical representations of the geographic name fields in the image records. Researchers are able to view this map, highlight any place or group of places, drilling down as far as street names and numbers, and then viewing the resulting set of records and images. In CHO Phase Two, searches in the GeoLocator system will be further refined and some of the problems with historical place names will be resolved.

Descriptive Records

International library and information search and retrieval standards (USMARC, Dublin Core, AAT, LCSH) will be used in creating descriptive records. Project participants will create high-quality end products that can be searched by a multiplicity of users and can be adapted to local institutional needs in OPACs and in-house inventories, guides, and finding aids.

Dublin Core was originally selected as the cataloging standard for CHO Phase One. However, ImageServer cannot produce Dublin Core records, nor can it import or export Dublin Core records. Therefore, it was decided to map the Dublin Core fields to MARC fields. Dublin Core definitions and MARC format are used to create the catalog records. The heading of each field is the label used in the Voyager OPAC for that field, followed by parentheses with the Dublin Core field name. The Dublin Core definition for each field and the MARC field number and sub-field(s) both appear as headings in the cataloguing module. For detailed Dublin Core and MARC mapping information, see the Connecticut History Online Project Dublin Core/MARC Crosswalk. The MARC field 042 |a dc is used to indicate that the MARC 21 record is derived from Dublin Core style record. It is hoped that in the future, if desired, these records can easily be converted to Dublin Core records.

Subject access requires the use of nationally sanctioned thesauri. For this purpose, the Art and Architecture Thesaurus (AAT) and the Library of Congress Subject Headings (LCSH) will be used. Many basic LCSH and AAT terms that were identified as important for CHO records during Phase One are included in the online Cataloguer's Manual that is part of the CHO website; this will be the primary source for subject terms to be applied to new material in Phase Two. AAT is available freely online and allows for maximum efficiency both in assignment of subject headings and in the retrieval of information. It will be used for terms relating specifically to the fine arts, decorative arts and architecture. LCSH will be available in hardcopy at the five participating institutions and will be used for other subject terms

Project staff will emphasize multiple qualities of the materials being cataloged in their subject heading analysis. For example, a portrait photograph's primary use is to document the appearance of the sitter; multiple secondary uses include documenting the style and type of clothing worn; the age, gender, occupation, and social status of the sitter; the circumstances under which the image was made (is it a studio photograph?); and the objects that are included in the photograph. Thorough subject analysis and the use of a consistent vocabulary of subject terms ultimately will allow users to retrieve different types of materials in a single search: for example, portrait photographs and actual costumes, or photographs of businesses and birds-eye views in which these businesses appear. Based on experience gained during Phase One of CHO, staff at both the CHS and the Dodd Center estimated the time required to catalogue the new types of material proposed for inclusion in Phase Two. The results varied, depending on material type and previously created catalog records, ranging from 15-20 minutes per record for photographs (to be added by the new partners, the New Haven Colony Historical Society and the Connecticut State Library) to approximately 45 minutes per record for maps and birds-eye views, since the latter will require extensive analysis and detailed subject headings in order to provide full access to their content. It is anticipated that oral histories will also require approximately 30-45 minutes per record.

Taking into consideration these factors as well as training time, vacation and sick leave, it was determined that full-time project employees will be able to produce between 1,500 and 2,000 descriptive records during a 12 month period for a total of 8,500 to 9,000 records from the five institutions. The cataloger at the Connecticut State Library will be engaged for a period of 18 months in order to create at least 2,500 records for objects in this important collection.

The cataloging process is expected to follow this pattern:

• The Project Coordinator will train the four new Catalogers (one based at each of the four participating institutions), and work with catalogers already at the University of Connecticut.
• The Catalogers will compile as much information as possible from the original item or inventory record and record the batch number and the batch item number. As part of the descriptive process, Catalogers will fill in a field that identifies the specific institution and assigns the institution's unique identifying number. These numbers will serve as links to the image on the Photo-CD and on the server.
• Catalogers will apply appropriate subject analysis and create narrative descriptions of each item incorporating all significant keywords. Subject headings will be applied using LCSH and. Personal and corporate names will be verified against the Library of Congress Name Authority File via OCLC.
• Links to the Map and Geographic Information Center database ( MAGIC) will be made from the cataloging record by the Catalogers using MapQuest (Tiger) and GINS for authority control.
• Descriptive records will be examined by the Project Coordinator on a regular basis to assure correctness, uniformity, and accuracy.
• During the last quarter of the project, each of the participating institutions will contribute to the OCLC database an average of 12 collection-level records of the materials that have been digitized.

Creating the Digital Copies

The project will continue to follow the pattern of Phase One and outsource the digitization work to corporate vendor. All of the partners would prefer to maintain their collections on-site. At the same time, they want to avoid the duplication of equipment and expertise. For preservation, insurance, administrative, and security reasons, it will not possible to send large quantities of original material out of the buildings to be copied. In addition, direct scanning of the original artifacts does not lend itself to the capture of a large number of images in a short period of time. Expert evaluation and adjustment of scanning resolutions, and image size slow the process considerably. On-site imaging also fell outside the stated goal of the project to provide an easily replicated model for other libraries and museums. For these reasons, it was decided to contract image capture to a private company with a good record of success. The project had good success working with Boston Photo Imaging (who has also done work for Harvard and New York University) and will continue to work with them in Phase Two to create both the film and the digital images for collection materials. (For additional information about Boston Photo Imaging please see Attachment III.)

Boston Photo Imaging will spend approximately one week at each institution to create film negatives of each standard size project image (up to 11 X 14 inches). The inter-negatives will then be converted to digital format in the form of Photo-CDs. For a small number of images larger than 11 X 14, materials from all institutions will be assembled at Mystic Seaport for copying. This process facilitates the capturing of large quantities of images without having to stop and make the constant adjustments that direct scanning would require. Each image will be recorded on the Photo-CDs at four resolution levels. For the purposes of public access, and to ensure image security, only two levels of image resolution will be available on Connecticut History Online: a low-level thumbnail image and a second mid-range resolution image.

Converting the images to digital format will require the following steps:

• Materials will be grouped by size in batches of 100 items before filming begins at each institution. “Batch sheets” will be created with the item accession number in the same order in which the items are filmed.
• The photographers film each batch in order on site.
• Film is developed and reviewed in Boston. Some items may be re-filmed and the new image inserted in its proper order.
• Boston Photo Imaging provides five different file sizes and resolution levels for each image on a PCD/FPX Kodak Photo-CD. Each Photo-CD represents one batch of images.
• The Project Coordinator reviews each image on the CDs to ascertain that it is of acceptable quality.
• After all photography is completed and all Photo-CDs have been created, each institution will create derivative JPEG files for placement on UConn’s server.

Boston Photo Imaging’s visits will be scheduled towards the beginning of the project. Since only a portion of the graphics at each institution will have actually been fully described by then, an accession or control number will have been assigned to every graphic so that photography can proceed.

Database Creation

The project will continue its existing website and augment this as the project grows. Copies of the new grant proposal, and technical information connected with Phase Two of the project will be added if the project is funded. Project and institutional staff will further review hardware and software configurations prior to placing orders for equipment. Consideration for final selection will involve current technology, hardware and software already installed and available at UConn, and the technical specifications of the project.

Connecticut History Online will continue to reside on a server at the UConn Libraries and will utilize the Endeavor Voyager OPAC interface, the Endeavor Voyager cataloging module, and the Endeavor Voyager ImageServer system . The Voyager OPAC interface, Voyager cataloging module, and the Voyager ImageServer module will enable the automatic production of USMARC records from descriptive cataloging records of the collection materials as they are being created. ImageServer is designed to connect with and deliver not only images but also manage audio, video, and text files. As a result, access to the metadata for all types of material from Connecticut History Online will be simultaneously provided to the end user via the Internet with immediate access the catalog record and the digitized item. Where institutions are not Endeavor sites, cataloging clients and ImageServer clients will be installed on individual personal computers for input of cataloging data. UConn has agreed to provide technical support to project participants and will maintain Connecticut History Online through their Information Technology Services Department.

Preservation

In addition to enhancing access to, and use of, visual collections this project will significantly contribute to the participating institutions’ efforts to preserve their irreplaceable collections. Having a significant amount of material online eases the level of use on collections and extends the life of original materials. The creation of digital surrogates for the 9,000 items included in this project will significantly reduce the need to consult the original artifacts. The CHO database itself will be preserved following the University of Connecticut’s standards for preservation of digital materials.

Back to top

MANAGEMENT PLAN

The project will cover the 24 month period from December 2002 through November 2004. It has three goals: expanding Connecticut History Online to capture an expanded time period and wider range of material, creating metadata that reflects the different types of material being digitized and developing a plan support the project into the future. Management of the project will reside with the Project Co-Directors, a steering committee representing the participating institutions: Nancy Milnor, Library Director, from the CHS; Mary Anne Stets, Curator of Photography, from Mystic Seaport; Thomas Wilsted, Director, the Dodd Center, Peter Lamotte, Director of the New Haven Colony Historical Society, and Kendall Wiggin, Connecticut State Library. The committee will meet twice a month for the first four months of the project and then once a month thereafter. Working under the supervision of the Project Co-Directors will be three distinct teams (Selection, Cataloging, Web design) headed by the Project Coordinator. The teams will be guided by the goals and timelines set out in the grant proposal and will use these as benchmarks to establish the viability and success of the project. Communication will come through regular meetings that rotate geographically among the partners and through e-mail and listserve communication.

On November 1, 2002, recruitment will get underway for the full time Project Coordinator. Recruitment of the four full time Catalogers, one of whom will be assigned to each of the participating institutions (except UConn), will begin in January 2003. The Project Coordinator will work on the project for 20 months and the Catalogers will work for 12 months at the three of the partners and 18 months at the Connecticut State Library because of the larger material being cataloged. Beginning work in March 2003, the Project Coordinator will: familiarize himself/herself with the five institutions’ personnel, procedures, and collections; participate in the hiring of the Catalogers; review specifications for and then order software and hardware; and confirm scheduling for Boston Photo Imaging’s work. Once the Catalogers begin work in April and July 2003, the Project Coordinator will supervise orientation and training of new staff and assume responsibility for the day-to-day management of the digitization component of the project. After initial training, the Catalogers will devote most of their time to providing item-level descriptions of collection material.

The Project Coordinator will be based at the University of Connecticut and will oversee the Catalogers to assure quality control, timely and efficient storage of project data, documentation of project activities, and the collection of statistics and management data. The Project Coordinator will also be responsible for authority control oversight. Quality control will involve spot-checking cataloging records for accuracy and completeness. The Project Coordinator will be present at each of the five institutions during the work of Boston Photo Imaging and will scrutinize Photo-CDs as they are generated to guarantee that the images are assigned the proper accession/control numbers The Project Coordinator will also be responsible for developing all descriptive texts for the project Web site, coordinate the activities of the Web site designer, and will be available on an as-needed basis to trouble-shoot cataloging problems. The Project Coordinator will lead the Web development team and be responsible for compiling and maintaining statistics. Resumes of all permanent staff and position descriptions for project staff are included as attachments. Please see the Schedule of Completion as well as the Tabular Plan of Work for additional details with respect to the management plan.

The project leader during Phase One was The Connecticut Historical Society and during Phase Two, the project leader will be the Thomas J. Dodd Research Center at the University of Connecticut. The budget will be managed through the University's Office for Sponsored Programs who will track and create budgetary reports.

Back to top