Conclusion

This paper presents a review of the various metadata schemas, organizations, initiatives and retrieval applications in the area of image collections. The number of different standards available for images is large, and both the research and applications seem to be accelerating in this area. The fast-pace of progress seems to have the drawback of quick obsolescence of many initiatives and working groups.

The theoretical foundation for some of the difficulties in the analysis of the visual medium suggests that the plethora of schemas in specific domains will continue to be a reality. For this reason, the image-metadata community is wise to focus on interoperability technologies such as the Resource Description Framework (RDF), Extensible Markup Language (XML) and compatibility with minimal metadata sets such as Dublin Core. Due to the ambiguity and complexity of visual information, it seems unlikely that we will find a single ideal metadata scheme for visual materials of all kinds. The most obvious candidate for this job would be the MARC standard, but studies show that archives and special image collections are not using MARC for reasons of efficiency and expressivity.

I recommend further comparative studies between the various schemas and technologies that currently exist, especially with the aim of integrating them into a common framework such as the Open Archive Initiative. I also recommend paying attention to the International Standards Organization (ISO) for principles of development of metadata schema.