Information Technology Reference
In-Depth Information
Next, the technique selects representative polylines for each cluster, as
shown in Fig. 9.2(d). Our current implementation simply extracts a
polyline as the representative closest to the centre of a cluster in an n-
dimensional vector space. This is a good strategy because it selects
average polylines. Here, if one or more polylines in a cluster have already
been selected as the representative polylines of other clusters, the
technique does not select any new representative from the current cluster,
so we can reduce the total number of representative polylines.
Interactive visualization
The extended technique represents the time-varying data as coloured
polylines. It assigns colours (e.g. blue to rainy, green to cloudy) to the tags
of weather data, and draws the polyline in the assigned colours. If the
vertices of a segment of a polyline have different tags, it interpolates the
colours along the segments. The technique also features selective polyline
display based on the tags. Users can select particular tags so the technique
can draw only corresponding parts of the polylines. They can also select
particular tags to be filtered from the display of the polylines.
Initially our technique draws only the representative polylines. Our
current implementation generates several clustering results, with several
configurations of the grid and the clustering process. Smoothly replacing
the clustering results, our technique seamlessly displays several levels of
numbers of representatives.
The technique adjusts transparency of representative polylines in each
time step. Here, a representative polyline is actually selected as a
representative in several time steps, while it is not selected in other time
steps. For example, Fig. 9.3 shows that the polylines A, C, and D are
selected as representatives, while A and C are actually selected as
representative in the former two time steps, and A and D are selected in the
last time step. In such time steps, in which a polyline is feasible for
selection as a representative, the technique draws the polyline in relatively
lower transparency based on the number of original polylines in the
cluster. Meanwhile, it draws a highly transparent polyline in the other time
steps. Through this process, the technique can reduce visual clutter while
displaying consecutive polylines. Also, users can observe how many
original polylines are represented as the drawn polylines.
Search WWH ::
Custom Search