Facilitating web search with visualization and data mining techniques

Abstract

Although conventional Web search environments are very useful for well-structured, closed-ended questions, they become less efficient in dealing with ill-structured, open-ended questions. This paper presents design rationales and implementations of an alternative Web search environment called Vis-Search. The VisSearch environment adopts visualization techniques to reduce various cognitive loads of iterative Web searches. It also employs a data mining algorithm to utilize Web search results of others. This paper describes in detail how these techniques can be used to facilitate efficient Web searches for illstructured, open-ended questions, thereby fostering creation of a community of knowledge on the Web. © Springer-Verlag Berlin Heidelberg 2005.