Although textual materials and their subsequent retrieval can be judged and evaluated on what they contain, images do not necessarily possess that characteristic for all viewers. As expressed by Gombrich (1960), in many cases images invoke emotion. Emotion derived from vastly differentiated levels of experience (and knowledge) appears to impact how viewers see images. While many approaches have been investigated to improve image retrieval, most to this point have failed to incorporate a means by which these affective elements can assist the retrieval process for system users. Without some form of "community of user" feedback for the extraction of affective subjectivity engendered by visual stimuli the process for arriving at functional taxonomies and effective indexing terminology will continue to be challenged in the arena of image retrieval.
展开▼