File system

Where are the big directories ? Were there a lot of pages created recently ?Which are the popular webpages? Are there some labs more popular than others? Which areas are getting more popular? less popular? Are new pages more popular that old pages? Which old page are popular? What proportion of the pages are never used? seldom used? ....

This questions can be answered by the visualization menu.
As an example, suppose that you want to find the biggest recent files. Choose in the Size Menu the parameter size with a mapping algorithm exponential. Then go in the Colors menu and choose the parameter ctime and use the distribution algorithm and the hue component. The biger and pinker nodes are those you are searching.
You can also select in the search menu in advanced mode the two parameters with an and operator and a minimal value for each one. This will give a selection that you can save and explore in a new tree window.

Can you see different patterns in those files?

This can be done in two step.
The first step (nlog(n) time) consists in a rough identification of the patterns. After what, same patterns would appear with the same color.
If this level of details is not enough then, select the subtrees that you want to compare and run recursively the previous algorithm. At the end, if there is not a lot of node in the subtree then run the Zhang algorithm. The complexity is n3. Thus, this last algorithm will not run if the total number of nodes is greater than 500 (this upper limit was fixed by experimentation on EVAT). The result is a color mapping of the nodes that are associated within the two subtrees.
An other application of the fast heuristic is to predic isomorph subtrees in one tree. This happen for example when some directories are dupplicated. On this image, one can see a big directory which is repeated in the file logs_A of the contest but also some small repeated one (look at the blue color).