Problems with databases and the XML solution

Abstract

Databases of biographical material are now a common feature of historical research. Many historians have converted archive-derived data into electronic format to facilitate the storage, querying and presentation of the material. In some cases, these electronic databases have been made available online. However, as biographical databases have proliferated on the web, problems have emerged, many of them linked to the type of technology used by historians. Most history databases are implemented using unsophisticated tools like Microsoft Access. While these systems are adequate as data-storage facilities they impose certain limitations. In particular, they oblige historians to shoehorn historical data into pre-existing categories that are not custom built for the material in question. Nor do they easily accommodate variation in the density of information between individual records. This results in the generation of unfilled ‘white space’ so that database size and inflexibility increase even though many of the records remain sparsely populated. While commercially produced databases have query facilities these become more difficult to operate as the individual database becomes more complex. Merging is also problematic.