[The explanation of the demonstration movie]

This is the demonstration movie of the presented visual analytics tool.
We visualize the credit card transanction logs of 1 month (the same as 
logs used in section 4.1.1)
 
0:00 Select "day" for x-axis.

0:05 Apply attribute recommendation (r3) which shows the attributes
     which include remarkable items.

0:07 Choose "shop ID" for Y-axis then observe the result (explained as 
     Figure 5 in the paper).

0:21 Apply clustering-based filtering.

0:28 Show representatives of each cluster.

0:32 Find the cluster which has the increase in the last weekend of 
     the month (cluster 3).

0:36 Choose cluster 3 to visualize in detail.
     (Note that the color of the button of cluster 3 is relatively 
     dark because a lot of shops belong to cluster 3.)
     The user can observe common behaviors of the shops belonging 
     to the cluster 3.
     (We can find that there were many transactions on other weekends.)

0:50 Stop applying clustering-based filtering and apply sorting-based
     filtering. Choose (s2) and adjust the number of items to be drawn
     using the slider.

1:05 Apply outlier detection and draw only outliers, only non-outliers
     and all items assigning pink to the outliers.
     The user can observe the behaviors of both outliers and non-outliers
     (explained in section 4.1.1).

1:22 Stop applying outlier detection and adjust the number of items 
     to be drawn.

1:26 Apply filters by mouse click (automatically choose the item).
     (Also the users can input specific values to be filtered for each 
     attribute using text fields.)

1:30 The user can continue observation applying attribute recommendation
     again and choose other attributes for Y-axis...