Fragment Identification · Multimedia

Abstract Cultural institutions and museums have real-ized that annotations contribute valuable metadata for search and retrieval, which in turn can increase the vis-ibility of the digital items they expose via their digital library systems. By exploiting annotations created by others, visitors can discover content they wouldn’t have found otherwise, which implies that annotations must be accessible and processable for humans and machines. Currently, however, there exists no widely adopted anno-tation standard that goes beyond specific media types. Most institutions build their own in-house annotation so-lution and employ proprietary annotation models, which are not interoperable with those of other systems. As a result, annotation data are usually stored in closed data silos and visible and processable only within the scope of a certain annotation system. As the main contribution of this paper, we present the LEMO Annotation Frame-work. It (i) provides a uniform annotation model for multimedia contents and various types of annotations, (ii) can address fragments of various content-types in a uniform, interoperable manner, and (iii) pulls annota-tions out of closed data silos and makes them available as interoperable, dereferencable Web resources. With the LEMO Annotation Framework annotations become part of the Web and can be processed, linked, and referenced by other services. This in turn leads to even higher visi-bility and increases the potential value of annotations