Chapter 10: User Interfaces and Visualization
Contents
|
Modern Information Retrieval Chapter 10: User Interfaces and Visualization |
Contents |
Faced with a large set of text collections, how can a user choose which to begin with? One approach is to study an overview of the contents of the collections. An overview can show the topic domains represented within the collections, to help users select or eliminate sources from consideration. An overview can help users get started, directing them into general neighborhoods, after which they can navigate using more detailed descriptions. Shneiderman [#!shneiderman96!#] advocates an interaction model in which the user begins with an overview of the information to be worked with, then pans and zooms to find areas of potential interest, and then view details. The process is repeated as often as necessary.
Three types of overviews are discussed in this subsection. The first is display and navigation of large topical category hierarchies associated with the documents of a collection. The second is automatically derived overviews, usually created by unsupervised clustering techniques on the text of documents, that attempt to extract overall characterizing themes from collections. The third type of overview is that created by applying a variant of co-citation analysis on connections or links between different entities within a collection. Other kinds of overviews are possible, for example, showing graphical depictions of bookshelves or piles of books [#!rose93!#,#!yates96!#].