Modern Information Retrieval
Chapter 10: User Interfaces and Visualization


Contents

next up previous
Next: 4. Co-citation Clustering for Up: 2. Overviews Previous: 2. Automatically Derived Collection

    
3. Evaluations of Graphical Overviews

collection overviews!evaluations

Although intuitively appealing, graphical overviews of large document spaces have yet to be shown to be useful and understandable for users. In fact, evaluations that have been conducted so far provide negative evidence as to their usefulness. One study found that for non-expert users the results of clustering were difficult to use, and that graphical depictions (for example, representing clusters with circles and lines connecting documents) were much harder to use than textual representations (for example, showing titles and topical words, as in Scatter/Gather), because documents' contents are difficult to discern without actually reading some text [#!kleiboemer96!#].

Another recent study compared the Kohonen feature map overview representation on a browsing task to that of Yahoo! [#!hchen98!#]. For one of the tasks, subjects were asked to find an `interesting' Web page within the entertainment category of Yahoo! and of an organization of the same Web pages into a Kohonen map layout. The experiment varied whether subjects started in Yahoo! or in the graphical map. After completion of the browing task, subjects were asked to attempt to repeat the browse using the other tool. For the subjects that began with the Kohonen map visualization, 11 out of 15 found an interesting page within ten minutes. Eight of these were able to find the same page using Yahoo!. Of the subjects who started with Yahoo!, 14 out of 16 were able to find interesting home pages. However, only two of the 14 were able to find the page in the graphical map display! This is strong evidence against the navigability of the display and certainly suggests that the simple label view provided by Yahoo! is more useful. However, the map display may be more useful if the system is modified to tightly integrate querying with browsing.

The subjects did prefer some aspects of the map representation. In particular, some liked the ease of being able to jump from one area to another without having to back up as is required in Yahoo!, and some liked the fact that the maps have varying levels of granularity. The subjects disliked several aspects of the display. The experimenters found that some subjects expressed a desire for a visible hierarchical organization, others wanted an ability to zoom in on a subarea to get more detail, and some users disliked having to look through the entire map to find a theme, desiring an alphabetical ordering instead. Many found the single-term labels to be misleading, in part because they were ambiguous (one region called `BILL' was thought to correspond to a person's name rather than counting money).

The authors concluded that this interface is more appropriate for casual browsing than for search. In general, unsupervised thematic overviews are perhaps most useful for giving users a `gist' of the kinds of information that can be found within the document collection, but generally have not been shown to be helpful for use in the information access process.


next up previous
Next: 4. Co-citation Clustering for Up: 2. Overviews Previous: 2. Automatically Derived Collection


Modern Information Retrieval © Addison-Wesley-Longman Publishing co.
1999 Ricardo Baeza-Yates, Berthier Ribeiro-Neto