Mildred L. Iddins Special Collections
Literary Society Archives
|
|
Mildred L. Iddins Special CollectionsLiterary Society Archives |
|
|
| Main | Collections | About | Acknowledgments |
| |
Creating the Literary Society Digital ArchivesIntroduction Selection Metadata Scanning Interface Design Preservation Copyright Quality ControlIntroductionUnder the direction of Bruce Kocour, Director of Stephens-Burnett Library, a project to digitize some of the Special Collections at Carson-Newman College (C-N) was begun in January, 2004. Al Lang, College Archivist determined, in response to the requests and perceived interests of alumni, the project should begin with the materials from the College's four literary societies. These societies, nearly as old as the college itself, served as the primary social point of student life from 1851 to the early 1980's. However, what were once active societies meeting in elaborate social halls and conducted as literary societies, have in the past 20 years degenerated into small groups more like typical college fraternities/sororities, or have ceased to exist altogether. The four societies, two for women (the Calliopeans and the Hypatians) and two for men (the Columbians and the Philomatheans), occasionally held joint meetings and this collection includes programs from a several of those co-sponsored gatherings. The materials from these societies are a valuable record of the more than 150 years that Carson-Newman has been providing a liberal arts education in a Baptist environment. Additionally, the record of literary societies in the heart of Appalachia may be a unique resource for scholars interested in Appalachian studies. The prototype digitization project was undertaken as a class project for the course "Digitizing Research Collections" at the University of Pittsburgh in the Spring Semester, 2004. Ms. Elizabeth Shaw was the Instructor for the class, and the members of the C-N Digital Archives Development Team were Tara Cassidy, Melodi Goff, Stanley Michalski, Karen Shephard and Luke Vilelle. The generous cooperation of Al Lang allowed the materials to be temporarily stored in Pittsburgh for the duration of the Spring Semester. The goal of the class project was to produce a sample digital resource which would be evaluated by the faculty of the Stephens-Burnett Library for possible inclusion in the library's website. Additionally, it was anticipated that images produced for the project would be submitted to the Appalachian College Association's Digital Library of Appalachia. SelectionSelection for the Carson-Newman literary society material digitization project was a difficult task. Faced with materials in all sorts of shapes and formats, from hand-written pages to a meeting announcement in the shape of a woman, we tried to establish a selection process that emphasized adding value to the collection through digitization. Our first decision was to focus on the Hypatian Literary Society (HLS) activity in the pre-World War II era. For this period, we wanted to create a prototype that could be used for future digitization efforts by the Stephens-Burnett Memorial Library, as it attempts to make more of its archival collection available online. Selection must match the mission statement of the sponsoring institution, and C-N's stated duty of "connecting and preserving historical materials related to Carson-Newman College, the Appalachian region, and to the Baptists of East Tennessee and the world" is consistent with digitizing the literary society collections. Although digitizing an entire collection is always preferable to "cherry-picking," or picking particular objects for digitization, we had to take the second route and choose from among the vast array of objects because of limited time and resources. C-N will likely have to do the same when attempting to further our work. One of the first principles of selection is to choose objects with previously compiled and complete metadata. The only parts of the collection meeting this standard were the pages of literary society information contained in the C-N yearbooks, so we decided to digitize all yearbook pages, including pictures and member lists, related to the Hypatians. The majority of the collection, though, is a potpourri of reunion notices, event programs, banquet announcements, and the like. Because none of these items seemed to have any more intrinsic value than any of the others, we considered a randomized approach to choosing from among the many programs and announcements -- for example, picking every tenth item. Some experts argue that such a sampling is more truly representative of the collection, because it is not affected by subjective judgments. However, our group could not resist picking out particular items for digitization that we thought best represented the collection. The variety can be seen in the materials available on this site. We also wanted to digitize materials representing the broad spectrum of media associated with the collection. Thus, we have digitized the musical score of the HLS official song and the HLS constitution. However, one aspect of the collection is not represented in the materials we digitized. We decided not to scan any of the hundreds of hand-written pages of meeting minutes and financial records. In many cases, these volumes are in poor shape and the pages are difficult to read. Using optical character recognition on these pages would not have worked, and a re-keying project would require a tremendous amount of labor. In other words, we saw no way to add value to these records through digitization. Even if we had scanned a few pages and mounted them on the web, there would have been no way to place those particular records in context, and the online user would have had great difficulty trying to read them. Certainly, if C-N has the resources in the future to consider re-keying the records or creating a searchable database of names present in the records, these would be valuable options. But for now, visiting the archives is the best approach to studying these records. With these digitization choices, we hope to have added value to the collection of literary society materials and created additional users of the collection. These digital materials will eventually be associated with the Digital Library of Appalachia, and we would encourage Carson-Newman to look for collaborative opportunities with other DLA members to create aggregate online collections of the literary society materials. MetadataMetadata is an essential element of any digitization project, and the success and longevity of the project are dependent upon identifying, collecting, and preserving all the pertinent metadata. Carson-Newman College's Mildred L. Iddens Special Collections digitization project is intended to meet the local needs of the college and provide materials for the Appalachian College Association's Digital Library of Appalachia (ACA DLA). The ACA is a regional consortium of thirty-four colleges and universities, and is actively seeking greater participation from members in the DLA project. For the prototype of the digitization project proposed for the Special Collections at Carson-Newman College, metadata concerns included the requirements of the DLA as well as those of Carson-Newman. Although the DLA has a set of metadata elements based upon the Dublin Core, other metadata schemes were reviewed for the Carson-Newman prototype. The DLA metadata element set does capture the essential information for web-display of a digital library, but does not address some of the administrative concerns which should be a part of a metadata scheme. Because the DLA scheme is essentially the Dublin Core set, this shortcoming is not surprising. The fact that the DLA modifications focus mostly on descriptive metadata is also not surprising, since that type of metadata is the one most familiar to librarians. However, the needs of a digital library are not identical to those of a physical library, and these differences should be considered when planning a project and establishing a metadata scheme. After reviewing many established metadata standards, including EAD, METS/MODS, and TEI, a modification of the DLA metadata set using the Western States Digital Standards Group (WSDSG) metadata scheme was selected. The strengths of the WSDSG complemented the weaknesses of the DLA. In most cases, the modifications employed were simply expansions on a DLA element using WSDSG elements. Because both standards are based upon the Dublin Core, it was anticipated that an amalgamation of the two would be relatively easy to devise. Unfortunately, that was not the case. Because the WSDSG addresses concerns the DLA does not, there are many cases of very confusing polysemous word usage. In the final version, the DLA name was preferred over the WSDSG designation and new terms were chosen, if necessary, for the metadata elements missing in the DLA. Local needs were also considered, and in instances where the DLA terminology seemed less than clear, modifications were made. Although these adaptations violate the desirable goal of a universal standard, it seemed more desirable to have clarity and a structure suited for the local collection. It should be noted that the DLA anticipated these types of concerns: "Each institution may decide to add optional metadata elements as their resources allow and as their catalogers demand" (Parker). There was no metadata recorded in any form for the material going into the project. Filling this void by collecting the basic descriptive metadata was a significant and time-consuming task, and is not fully completed since the Archivist at Carson-Newman will need to be consulted for some details. For the prototype digital project 52 images of 28 unique objects were created, and metadata about those images was also recorded. The metadata will be entered into a Microsoft Access database developed for this prototype project. Because of the nature of the metadata and the many possible relationships between items and types of metadata, a relational database is the ideal location for metadata storage. Also, although the prototype is a static website, it is anticipated that the site will grow and that searching will be incorporated into the design. A relational database will be crucial to this new feature. Unfortunately, creating the database proved more challenging than anticipated, and for the sake of allowing the project to move ahead, the metadata was collected in written form as an intermediary step. This is definitely not an ideal approach. Entering the data into the database is going to be a time-consuming task, and there is likely to be some difficulty in reading the handwritten notes. Nevertheless, by turning to "the old fashioned way," the information was gathered and progress was made in other areas. Perhaps more importantly, as a result of seeing what type of data was entered by hand for the different elements, the database designer was able to identify field names that were likely to be problematic and make changes or notes as needed. A crosswalk from DLA to C-N is provided below, along with the location of the element in the Access database.
ScanningDigital imaging for this project began with a thorough review of current imaging standards, developing imaging criteria for this project then selecting materials from the collection for scanning. A single Master image file for archiving purposes was prepared for each selected item. The Master image was prepared by scanning the item and saving the scan in the loss-less Tag Image Format File or tiff format. From the Master image two derivatives were made, one for high quality display purposes and one for a clickable thumbnail. The derivative images were saved in the Joint Photographic Experts Group or jpeg format for display purposes. The process began by reviewing current imaging standards then producing a test of our criteria by scanning six items from the collection and preparing derivative images for evaluation of overall resolution and quality. Selected materials were scanned on a Hewlett-Packard flatbed scanner in 24-bit color, loss-less compression at 300-dpi capture resolution for producing the Master tiff image file. Before saving, the images were slightly cropped, and contrast adjusted for best visual on-screen appearance. No other adjustments were made to the Master image. Even though most of our materials are in black and white, half tone or text, we decided to scan all materials in 24-bit color for improved richness of tone Derivative images were produced from the Master image with Photoshop 6.0 software. Each image was manipulated to varying degrees for acceptable presentation. Many images required rotation, additional contrast adjustment and cropping, sharpening and/or blurring (for halftone pictures). One image was oversize and required stitching two images together. The display and thumbnail images were saved in jpeg format. Following all adjustments in Photoshop, each image was viewed in Photoshop's Image Ready software for selection of best image quality and file size combination. The Image Ready optimization software allows for simultaneous viewing of three image variations and the Master image for visual comparison purposes. An attempt was made to maintain display image file size to around 50k or less for fast loading. The thumbnail derivative incorporated the display image adjustments and was made immediately following saving the display image. Since most images were oriented vertically and some were square, the thumbnails were all uniformly sized across the top width of the image, optimized in Image Ready, and saved. Fifty-six display images plus fifty-six thumbnails were prepared with a combined total file size of approximately 2.6 MB. The Master image files and all derivatives were saved on two CD-R's. Quality control was maintained by conducting the scanning work at one time with two persons. A scanning and record keeping procedure was developed for the session. The same scanner settings were used throughout and final archive image quality was based on the visual acceptance of both persons. InterfaceThis site was designed using Dreamweaver and cascading style sheets to enforce a standard appearance. The guiding ideas for design were that the look of the site and the navigation scheme should be clean, logical, and consistent. The total materials pertaining to a single society constitute a collection. Within each collection, content is organized by category, such as yearbook photos or artwork. Collection submenus present each category's content as a list of thumbnail images. Each thumbnail is linked to the item display page, which opens in a new window and shows the full size image with available identifying information. The site uses two standard page designs: navigation and display. The navigation page includes the global navigation panel with links to each area of the site:
The global navigation panel highlights the user's current position to avoid confusion. The navigation design is also used to navigate specific society collections with the addition of collection submenu options (Yearbooks, Artwork, programs, etc.) that lead the user to access the display pages. When a user enters a collection submenu, the name of the society appears at the top of the submenu to facilitate movement within collections. The display design is used to present the full size JPEG images, and appears as a new window. This page type lacks navigational panels for two reasons: to enhance the aesthetics of the image display and enable the user to view more than one image at a time, since the user is responsible for closing the display windows instead of being required to do so in order to continue browsing the site. PreservationPreservation is a two-fold issue for the Carson-Newman Literary Societies Collection. Measures to protect and preserve the original materials and measures to ensure the ongoing maintenance of this digital collection must be addressed. Since a main goal of the digitization project is to generate more interest in Carson-Newman College, a plan to protect original items is essential as alumni and other interested individuals may want to access them. Many of these materials are in a very fragile state; therefore, decisions must be made concerning the handling of the items. It may be preferable to have viewing access only to some of the most delicate items, or copies of these items may be made for users to handle. A flyer discussing the proper way to handle collection items should be given to and read by all visitors. Cotton gloves should be used by everyone touching collection items, and a staff member should monitor visitors to ensure proper handling and security. Regular inventory of the collection is essential to ensure that all items are present, as well as to note any changes in their condition (i.e. pencil marks, tears). Where the collection is housed is critical. It is most important for the collection to be house in an environmentally safe room, preferably between 65 and 75 degrees, with a relative humidity around 50 to 70 percent. Frequent fluctuations in temperature and humidity can harm the collection by affecting the composition of the items. Too much humidity can result in expansion, while too much heat or too little humidity may cause items to become brittle. An up-to-date HVAC system can greatly improve environmental conditions. Lighting is another concern. Collections should never be exposed to direct sunlight and artificial lighting levels should be kept low. Preservation-approved cases and shelving units should be used to house materials, since the chemicals in many items may react to those in non-approved units. Also, like item types, photographs especially, should be kept in individual envelopes with like items of the same size, to reduce impressions caused when multi-sized items are stored together. The Northeast Document Conservation Center (NEDCC) web site is a great resource for preservation issues. It also offers an excellent handbook on digital preservation entitled the Handbook for Digital Projects: A Management Tool for Preservation and Access Preserving, or taking good care of original items during the digitization process a concern. Decisions must be made regarding the best method to use to digitize an item. Perhaps the item is too fragile to digitize. Some bound items may be too brittle to open for a flat scanner; however, scanning on a book cradle may work. Still, to protect the integrity of some items, it may be necessary to scan only the outside cover and manually enter the text, or a summary of the contents, leaving the original item intact. Once the items have been digitized, the digital collection itself becomes a preservation issue. To ensure that a quality collection is maintained over the years, standardization is essential. When digitization is done following standard methods, it is more probable that in future years migration to a new technology will be possible. Backing up the collection on CD-ROM or magnetic tape is also a preservation essential. And both the online collection and the backups should be checked routinely to determine if a problem exists (i.e. a link has been altered). Maintenance of both hardware and software is also critical to the preservation of digital collection. And metadata on the creation and storage of data, file format, intellectual property rights and their management, and software and hardware platforms will enable reconstruction of information, if needed in the future. CopyrightCarson-Newman College is the owner and copyright holder of the materials in the Literary Societies Collection. Digitization of these materials complies with United States Copyright Law (Title 17, United States Code). The copyright law does enable users, under the "fair use" section, to make a photocopy or other reproduction of an item if it is for the purpose of private study, scholarship, or research. For copyright permission regarding other uses, please contact Stephens-Burnett Library at Carson-Newman College. For more information on copyright law, see Stanford University Libraries' Copyright & Fair Use web site or view the complete version of the U.S. Copyright Law at the Library of Congress web site. Quality ControlThroughout our digitization project quality control was implemented repeatedly. With nearly every undertaking, Development Team members worked together, checking one another's work. When one segment of the project was completed, it was presented to the group, which then examined it, detecting and correcting problems such as coding or typographical errors, inconsistencies, dead-end links, and so forth. Scanned tiff images were re-examined when derivatives were made, and derivatives were examined when they were mounted to the web site. As different sections of the site were created, group members were informed and comments and suggestions were encouraged. Links were double-checked and metadata was reviewed. This checks-and-balances system enabled us to catch problems early in the process and correct them. By re-examining our work and sharing input, we also recognized where improvements could be made immediately, or considered for the future. A plan for quality control is essential for every digitization project. We recommend that Carson-Newman's digitization team include the following in their quality control plan: Quality Control Recommendations
Many of the above recommendations were gleaned from the 2000-2003 Moving Theory Into Practice: Digital Imaging Tutorial, Chapter 4, "Quality Control." When a conscious effort is made to review elements of the project as they are developed or created, problems can be recognized and corrected, avoiding time-consuming, costly, and sometimes irreversible errors. Source Cited: Fulton, Wayne. Moving Theory Into Practice: Digital Imaging Tutorial Parker, Kathy and Robyn Williams
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Albert L. Lang, Archivist
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||