Applications of a Lightweight, Web-Based Retrieval, Clustering, and Visualisation Framework

摘要: Today's web search engines return very large result sets for query formulations consisting of few specific keywords. Results are presented as ranked lists containing textual description found items. Such representations do not allow identification topical clusters, and consequentially make it difficult users to refine queries efficiently.In this paper, we present WebRat, a framework web-based retrieval, clustering visualisation which enables parallel querying multiple engines, merging retrieved sets, automatic clusters interactive the refinement. This is lightweight in sense that consists small, platform-independent component can be easily integrated into exisiting Internet or Intranet forms without requiring system environments, server resources precalculation efforts.The WebRat extends existing approaches many aspects: Found results added incrementally they arrive, labelling performed 2-dimensional space on user see rendering optimised provide sufficient performance standard office machines.The has been used implement variety applications: We have provided enhanced capabilities doing scientific research. Overview refinement implemented environmental domain. Finally, abstracts generated fly by knowledge management navigation developers searching technical information mailing list archives.