Figure 1 is a simplified illustration of a map, which contains two major clusters
of nodes with high rating, A and B, and an outlying node P. The clusters
indicate that most of the user's interests fall into two categories. If the user is
interested in one of these categories more than the other, he or she can browse
the recommended items around the cluster. If the user wants something different
from the recommendation, the area around the node P can be a starting point for
exploration. In a statistical viewpoint, P is regarded as an outlier to be excluded.
In the user's viewpoint, however, the fact that this node lies far from the clusters
can be a clue to a new interesting category.
|Other users' comments on items are presented with the commentators'
tastes visualized as the color patterns of their ratings.|an item with high rating a recommended item a less recommendable item Figure 1. Color pattern of a map
4 Highlighting by SimilarityThe layout of the nodes gives only approximated information on the actual
semantic distances. To represent more detailed relationship between the nodes,
we introduced a dynamic filter that highlights nodes similar to specified nodes.
The user can browse details of the similarity space by selecting nodes and
controlling the threshold of similarity:
|• ||By selecting a known item (i.e., items with the user's rating), the user can
dynamically extract similar items from the map.|
Questia, a part of Gale, Cengage Learning. www.questia.com
Book title: Human-Computer Interaction:Communication, Cooperation, and Application Design.
Contributors: Hans-Jörg Bullinger - Editor, Jürgen Ziegler - Editor.
Publisher: Lawrence Erlbaum Associates.
Place of publication: Mahwah, NJ.
Publication year: 1999.
Page number: 134.
This material is protected by copyright and, with the exception of fair use, may
not be further copied, distributed or transmitted in any form or by any means.