Large numbers of maps are accessioned regularly by map libraries, where they are a major source of information for geographers and other scholars. The process of cataloging those maps is time-consuming, expensive, and prone to error, however, because many catalogers have little substantive knowledge about the information contained on the maps. The complexity of information presented on many maps likely will prevent the complete "automation" of the map cataloging process, but partial assistance is possible through development of semi-automatic systems that can be use to rapidly classify much of the information on maps. This doctoral dissertation research project will conduct a more comprehensive inventory of the criteria currently used to catalog maps. These criteria then will be used to define a set of decision rules and develop a computer-based program to assist nonexperts in classifying maps, and that program will be tested in an experimental framework to determine if it is faster and more effective than current methods. The first part of this project will increase general theoretical understandings about how map information is used to catalog maps, while the second and third offer prospects for development of a minor "expert system" that can economically assist in the cataloging process. This project should help determine the degree to which certain steps in the map cataloging process can be efficiently handled by nonexperts, thereby freeing persons more knowledgeable in the content of the maps for more sophisticated analyses.
