Building a Semantic Web Image Repository for Biological Research Images

Abstract. Images play a vital role in scientific studies. An image repository would become a costly and meaningless data graveyard without descriptive metadata. We adapted EPrints, a conventional repository software system, to create a biological research image repository for a local research group, in order to publish images with structured metadata with a minimum of development effort. However, in its native installation, this repository cannot easily be linked with information from third parties, and the user interface has limited flexibility. We address these two limitations by providing Semantic Web access to the contents of this image repository, causing the image metadata to become programmatically accessible through a SPARQL endpoint and enabling the images and their metadata to be presented in more flexible faceted browsers, jSpace and Exhibit. We show the feasibility of publishing image metadata on the Semantic Web using existing tools, and examine the inadequacies of the Sem...

• Type: Conference
• Year: 2008
• Where: ESWS
• Authors: Jun Zhao, Graham Klyne, David M. Shotton
• Keywords: ESWS 2008, Image, Image Repository, Internet Technology, Semantic Web