Abstract
Many Japanese public libraries provide their resources on the Web. This paper shows an analysis of about 5,000 resources collected from the Web sites of 121 public libraries in Japan. We first collected resources from the library Web sites and created a metadata for every single resource. Based on the analysis on these metadata, we created a metadata for each of 194 collective resources (collection metadata). The metadata schemas were defined based on Dublin Core. We analyzed subjects and resource types using the collection metadata. We found that the major subjects were regional history and culture, especially after the Edo era. On the other hand, academic resources in natural and social sciences were not dominant. We also found that many libraries provided their resources on the Web but not all of the resources were fully digitized. The libraries provide lists of their resources and/or partially digitized resources with their descriptions.
