The Visual Analysis System (VAS) was developed to couple emerging successes in data mining with information visualization techniques in order to create a richly interactive environment for information retrieval from the World Wide Web. VAS's retrieval strategy operates by first using a conventional search engine to form a core set of retrieved documents. This core set is expanded by crawling from links in these documents to form a much larger set. The link topology of the larger set is then examined using data mining techniques to identify pages likely to be both relevant to the query and reliable as sources of information. Information visualization techniques are used to display the filtered set in a form amenable to the use of perceptual processes to filter and guide further user directed search. (Contains 16 references.) (MES)