GUIDO: the Visualization of Document Space

With increased emphasis on information retrieval from large, full-text document databases, and interest in browsing such databases, there has come an interest in improved techniques for visualizing the organization of documents within the database. In this paper we present a document set visualization based on a vector space model, with four significant features. First, the approach allows the use of multiple reference points in contrast to a single query or user profile. Second,the basis of the visualization is the distant of each document from the each of the reference points. Third, the visualization display large numbers of documents, encouraging browsing. Fourth, the visualization is highly flexible, allowing users to select among various reference points, distance metrics, and retrieval strategies. the visualization is implemented in a prototype system called GUIDO (Graphical User Interface for Data Organization).