CATC

Clue-aware trajectory clustering

CATC exploits clues betwwn trajectories to group similar trajectories into clusters such than each cluster represents one movement behavior of the user.

CATC based on CATS values and consists thress phases:

  1. clue-graph generation phase
    • A graph structure is used to represent CATS values among every pair of trajectories
  2. core set identification phase
    • Trajectories that have large CATS values to each other as core sets.
  3. cluster discovery phase
    • Core sets are merged as cluster, each cluster represents one movement behavior

results matching ""

    No results matching ""