An Integrated Space-Time Pattern Classification Approach for Individuals´ Travel Trajectories

One significant challenge for scientists is how to mine useful patterns of moving objects´ trajectories, with the increasing individual data collected by location-aware technologies. This paper proposes an integrated space-time pattern classification approach for individuals´ travel trajectories, which differentiates itself from traditional data mining techniques, such as clustering, frequent pattern discovery and so on. This approach can classify these trajectories by virtue of taking movement´s direction, distance, and time into account, and has the advantages over traditional data mining techniques in the aspect of space-time pattern mining. The experimental results has demonstrated its ability of supporting space-time pattern analysis and its capability of classification for a huge amount of trajectories.