This paper seeks to report on the study that proposed a model of image retrieval tasks for creative multimedia. The aim of this model was to understand the purpose of the tasks, intended use of the images, mode of query submission, nature of the keywords given by the users, and relevance criteria.
A survey was done to compile a total of 35 image retrieval tasks from 35 academic staff members at Faculty of ICT, International Islamic University Malaysia, and Faculty of Creative Multimedia, Malaysia. A search using Google Image Search category was carried out to find images on the web that met the intended use of the academic staff members.
Findings revealed that images were mostly intended for analysis, decorations, design, illustrations, image processing, and inspiration. Users preferred linguistic query mode, and visual query mode if they had a sample of the image. Most users requested images with captions for making the relevance judgment. Technical attributes, topicality, and completeness were the most important relevance criteria. Users' keywords were of abstract and concrete elements, and were expressed in a visual way and as a subject. Images decided as relevant ranged from an object in the image to the whole image. This model reflected similar findings to other studies with some variations.
Results are useful for understanding the nature of image retrieval tasks for the area of creative multimedia.
This paper developed a model of image retrieval tasks in the area of creative multimedia, and offers a value in understanding the tasks that are intended to meet the demand of the Creative Multimedia Industry established in Malaysia.
CitationDownload as .RIS
Emerald Group Publishing Limited
Copyright © 2005, Emerald Group Publishing Limited