Topic cascades: An interactive interface for exploration of clustered web search results based on the SVG standard

Abstract

The WebRat is a light-weight, web-based retrieval, clustering and visualisation framework which can be used to quickly design and implement search solutions for a wide area of application domains. We have employed this framework to create a web meta search engine combined with an interactive visualisation and navigation toolkit. Based on the SVG graphics standard, this application allows users to explore search results m a quick and efficient way, by choosing from topically organized result groups. The visualisation and the cluster representation can be stored and reused. We have combined hierarchical navigation of search result sets with a topical similarity based arrangement of these results in one consistent, standard-based system which demonstrates the potential of SVG for web-based visualisation solutions.