In this preliminary research, we discuss techniques to improve the quality of image retrieval and image management with the help of context information over the web. Our hypothesis is that leveraging the semantic annotated contextual metadata of the image would yield the relevant search results and facilitate building a consistent, unambiguous image knowledge base. Inferencing capability of semantic technology is very useful to deduce the new metadata from the existing metadata about images. We discuss the three different aspects of image context such as spatial, temporal and social context. Spatial and temporal context of the image can easily be gathered by taking advantage of modern digital camera technology. In order to acquire social contextual metadata, we benefit from social network sites.
展开▼
