Executive  summaryChapter1Chapter2Chapter3


Chart 1: Unstructured Data Preparation

This chart illustrates the hours worked to make unstructured data available. During Year 1 only three universities, (American, Maryland and Michigan) processed unstructured data. All three of them spend the same amount of hours in Year 1 as in Year 2. Maryland spent the most time at 211 hours both years. In Year 2, Columbia and Virginia made unstructured data available; they spent 21 and 4 hours respectively. Cornell and Illinois did not make unstructured data available either year.


Procedures: Unstructured data was copied to the Web server, then it was hyperlinked from the structured data screen. All unstructured data was displayed as provided by the museums; the only changes made were the addition of background texture, and the addition of HTML tags to indicate title as opposed to body text.


Procedures: Unstructured data was made available as linked text-only (ASCII) files; no additional processing was done. References to the unstructured data were provided in the structured records, and the exporting program linked the files. A programmer wrote code in shell script.


Procedures: Unstructured data was made available through a feature in SearchSlide™. After the unstructured files were placed in their subdirectory, the SearchSlide™ program determined the path using the file names provided by the museums and the filename contained in the record field. The data was uploaded from the CD-ROMs without any processing. There were no problems encountered during this process. Programs were written in Visual Basic


Procedures: Perl script was used to generate all the detailed information displays in advance, then links were made by the script to associated text files. In order to deal with the inconsistencies across the museum data, they wrote separate Perl scripts for each museum. The data was made available through a hyperlink from the full record. They added HTML headers and footers and often converted the body of the text to HTML in order to enhance the formatting.


Procedures: Unstructured data was made available during Year 2. A simple link was made by Perl script, which wrote the full record HTML page to check for the existence of an unstructured text and make a link to the path/filename if there was one.

Table 2: Universities and the skills needed to develop database systems

Back to Chapter 3


The Cost of Digital Image Distribution:
The Social and Economic Implications of
the Production, Distribution, and Usage of Image Data

By Howard Besser & Robert Yamashita