Student Equity, Excellence, and Diversity

Digitizing Hawaiian Language Newspapers to Promote
Wider Use via the World Wide Web

Developments Related to a Large-Scale Continuation of Project

1. Off-site, scheduling problems
   Coordinating the scheduling of the rented equipment, which was located off-campus with the work schedules of students from different campuses required intense and continued communication between all parties concerned. We lacked back up equipment in sufficient quantities, so equipment had to be transported back and forth at very specific times to ensure complete backing up of scanned image files.
2. Equipment, software problems
   Post-scanning processing, problems and procedures took up 95% of the Systems Librarian's time in the project. The following are examples of problems and issues encountered:
• The Minolta scanning system did not allow any flexibility in naming of files or saving of files (either location on the hard drive or format of the graphic file). This caused extra complexity when trying to identify image files after they were transferred from the leased off-site equipment to the microcomputer equipment at Hamilton Library, where post scanning work was conducted.
• In addition, we encountered problems in locating a graphic manuipulation program that would perform all the necessary processing functions. Several of the graphics programs which we explored would not open the files in the only format in which they could be saved on the Minolta scanning system. To use the PhotoShop or Canvas graphic programs, one had to first convert the images to a different file format using PaintShopPro.
• Overall limitations of the microcomputer hardware prevented the use of all of the software. For example, the OCR software (Cuneiform) ran slowly and required more memory than was available on the 386 micro which we used. Therefore it was not feasible to process the image files into ASCII text. To demonstrate the limitations of the 386 equipment, the SLIS digitizing intern working during the summer did time trials of the scanning cleanup and graphic software comparing the 386 versus a pentium microcomputer. Time requirements for various tasks were drastically longer for every task.
• We began by scanning Ka Hoku o ka Pakipika, a historically significant newspaper which was begun by King David Kalakaua. Ka Hoku o ka Pakipika images could not be scanned in one page/one scan mode using available lenses and scanner, so each page was scanned in sections. Problems with reattaching the sections into one image for the World Wide Web included:
1. the total size of the images was too large and loaded too slowly to be very usable,
2. inability to match the sections exactly,
3. inconsistent lighting in the scans.
• Use of the scan clean up software Scanfix was found to be unworkable, possibly because of the lighting problems noted above and partially due to the non-standard type faces and poor condition of the newspapers microfilmed. We learned that the only other institution scanning similarly aged newspapers from microfilm that also uses the Scanfix software employs it only to remove margin line marks, not to attempt to clean up the text images. Further testing of the Scanfix software will be done by the SLIS intern using images scanned from the paper printouts.
3. Further work
   We have found that this project of the University of Hawai'i at Mänoa, Hilo, and the Honolulu Community College Libraries is at the forefront in the creation of digital libraries that preserve our nation's heritage. We continue in our efforts to provide electronic access to primary newspaper archives and to provide for the preservation of and access to our cultural and scholarly heritage continues. We have learned that making source materials available in digital format requires investment of large amounts of time and labor. However, we feel the end product will be highly valued and contribute to cultural and historical understaning. We expect to present a summary of our experiences at a future conference such as the Digital Libraries '98: Third ACM Conference on Digital Libraries.

   A Pentium computer has recently been assigned to the project by the UH Mänoa Library administration. In addition, the Library's Desktop Network Services department plans to acquire a second Pentium computer and related digitizing equipment through other sources to increase the efficiency of our project. We are searching for grant opportunities which will allow us to continue the scanning/digitizing/processing for the Web of Hawaiian language newspapers. There are currently approximately 58 Hawaiian language newspapers, in approximately 93 reels, that await this work.

Back To Top