Abstract

Search result visualization is a task performed by search engines that enables users to find their desired documents, in an effective and efficient manner. Image based summary or best images of a web document, displayed as a part of the visualization process, has become indispensable, as a human perceives images instantaneously. But, selection of the best image increases latency in search result generation, and workload for the search process. In this paper, we propose and implement a search framework by integrating text and image search engines that increases the speed of extracting a representative image of a web document. Text associated with an image, image area and position are incorporated with the ranking function that finds the image snippet. By comparison, we show that our framework significantly improves over the existing ones in terms of time complexity, while maintaining the quality of image based summaries.

This work is licensed under a Creative Commons Attribution 4.0 License.