User #000

The tree of the user's mobility. It is composed of two prefix trees starting from a unique location: one encodes the paths arriving at it, the other the paths exiting from the root location.

A prefix tree is a particular structure that groups together the travels that share the same initial path.

One location can then appear in different branches because it can occur after different sequences of traversed points.

The trajectories entering the root (the central point) can continue and become outcoming paths.

The same representation of the mobility prefix tree but mapped on the distance travelled.

For each depth level of the incoming and the outcoming tree, it represents which percentage of all the leaves (start/end of a trip) are in that level.

For each km far from the incoming/outcoming tree, it represents which percentage of all the leaves (start/end of a trip) are in that level.

The median and the maximum number of different successors that a location node can have.

This information is useful to understand how fragmented is the user's mobility and how broad are the hubs (that is locations with many possible following directions).

The location richness of a tree is the proportion between the number of all the points traversed and the number of different locations encountered.

It gives information about how much regular are the user's movements: if the richness is low, it means that the same locations are traversed quite frequently; on the contrary, an elevated richness means that the user explores many different locations, and rarely returns on the same routes.

The distribution of the entropies of each location.

The entropy is the amount of bytes necessary to describe a location, according to Claude-Shannon's information theory.