Graphics Reference
In-Depth Information
Conclusion
8.5
he use of multiple linked views is a well-known concept in interactive statistical
graphics.Byestablishing acloseconnection betweenplotsthat showdifferentaspects
of related data, the analyst can explore the data in an easy and flexible way. Linking
simplelow-dimensionalviewsenables theusertounderstandstructuresandpatterns
of more complex datasets. Multiple linked views are an essential contribution to the
field of visual data mining and can provide the required human-computer interac-
tion (Fayyad et al.,
) to understand the hidden structures and relations. Embed-
ded in e
cient sotware systems, the paradigm of linked views is a powerful tool to
exploreabroadrangeofdata.hesimplicityoftheindividual plotscombinedwithan
intuitive, easy-to-use, and flexible user interface is especially rewarding when using
it for consulting experts in the data domains. Applied researchers are familiar with
most of the displays used as ingredients in linked systems. Hence a broad audience
can easily use and interpret these plots.
Linking procedures become particularly effective when datasets are complex, i.e.,
they are large (many observations) and/or high-dimensional (many variables), con-
sist of amixture of categorical andcontinuous variables, and have alot of incomplete
observations(missingvalues).Generalizationoflinkingaimstogiveconsistentviews
of data, consistent not only for each individual point but also for a plot as a whole. To
offer consistent comparisons of visual displays, plots should have the same scale and
should allow one to compare proportions. he interactive spirit of an analysis offers
a way to build in prior knowledge and metadata. Impressive results have been ob-
tained in the exploratory analysis of spatial data; the same can be expected for other
areas.
References
Anselin, L. (
). Interactive techniques and exploratory spatial data analysis, in
P. Longley, M. Goodchild, D. Maguire, and D. Rhind (eds), Geographical Infor-
mation Systems: Principles, Techniques, Management and Applications,Wiley,New
York, pp.
-
.
Becker,R.A.,Cleveland,W.S.andWilks,A.R.(
).Dynamicgraphicsfordataanal-
ysis, Statistical Science
:
-
.With discussion.
Diaconis, P.N. and Friedman, J.H. (
). M and N plots, in M. Riszi, J. Rustagi and
D.Siegmund (eds), Recent Advances in Statistics: Papers in Honor of Herman Cher-
noff on His Sixtieth Birthday, Academic Press, New York, pp.
-
.
Fayyad, U., Piatetsky-Shapiro, G. and Smyth, P. (
). he KDD process for extract-
ingusefulknowledgefromvolumesofdata,Communications of the ACM
:
-
.
Hofmann, H. (
). Graphical Tools for the Exploration of Multivariate Categorical
Data,BookonDemand.
Hofmann, H. and heus, M. (
). MANET, http://stats.math.uni-augsburg.de/
manet.
