Spatio-Temporal Data Mining and Classification of Ships' Trajectories
14 likes3,988 views
The document discusses spatio-temporal data mining and classification of ships' trajectories, focusing on knowledge discovery from trajectories, clustering methods, and similarity measurements in the context of maritime activity. It outlines methodologies including the use of fuzzy logic for trajectory qualification and visualization techniques for spatial data. Future work is highlighted, emphasizing enhancements in statistical analysis and real-time data processing.
Spatio-Temporal Data Mining and Classification of Ships' Trajectories
- 1. Spatio-Temporal Data Mining and Classification of Ships' Trajectories Laurent ETIENNE PhD in geomatics French Naval Academy Research Institute Geographic Information Systems Group Maritime Activity and Risk Investigation Network Department of Industrial Engineering, Dalhousie University [email protected] Halifax, June 2012
- 2. Introduction Movement is an important part of life Mobile objects tracking systems Large spatio-temporal databases Knowledge Discovery from movement Real time analysis Decision support systems Different kind of mobile objects Different mobility data interest Ecology, Sociology, Transports, Intelligence... 2
- 3. Research interests Knowledge discovery from moving objects databases (KDD) Algorithms for spatial data processing and modelling Advanced visualisation techniques for spatial data 3
- 4. Process overview 4
- 5. Spatio-temporal data mining Extract knowledge from a data warehouse Cluster groups of trajectories Main route followed by most trajectories of this group Main trajectory Spatial spreading (channel) Temporal stretching (channel) Metrics and rules to compare trajectories to main routes 5
- 6. Trajectories comparison Frechet distance and Dynamic Time Warping Frechet : Minimise the max distance between pos DTW : Minimise sum of distances between pos 6
- 7. Group of Similar Trajectories The model allows trajectories clustering using : Distance (Fréchet, DTW...) Density (T-OPTICS) Zone Graph (Itinerary) 7
- 8. Main trajectory Median trajectory Cluster positions (Normalized time, Frechet, DTW) Compute aggregated median position (K-Mean) 8
- 9. Statistical analysis Statistical analysis of points clusters distribution (distance, time, heading...) Boxplot visualisation 9
- 10. Spatio-temporal pattern Median trajectory and spatio-temporal channel Cluster positions (Frechet matching) with the main trajectory positions Compute spatial and temporal distance to the median position Sort spatialy (left/right) Sort temporaly (early/late) Statistical selection 90% Normality bounds ∆left / ∆right ∆early / ∆late 10
- 12. Qualify a Position Spatio-temporal channel Merge together spatial and temporal channel At each relative time of the median trajectory Normality bounds 5 zones defined Qualify a position How to qualify a trajectory ? 12
- 13. Similarity measurements Average, maximum and variability of spatial/temporal distance between the trajectory and the spatio-temporal channel (%) 13
- 14. Fuzzy Logic Spatio-temporal similarity classification of a trajectory compared to a pattern Using Fuzzy logic : Fuzzy sets learned by statistical analysis of similarity measurements Fuzzy rules defined by experts and combining similarity measurements 14
- 15. Fuzzy Logic (Fuzzy sets) Use statistics of similarity measurements Min 20% 40% 50% 60% 80% Max Define fuzzy sets 15
- 16. Fuzzy Logic (Fuzzification) Match a trajectory to the spatio-temporal pattern (Frechet matching) Compute the similarity measurements Fuzzify similarity measurements using fuzzy sets Value : 145 75% Medium 25% High 16
- 17. Fuzzy Logic (Fuzzy Rules) Apply fuzzy rules using a fuzzy associative matrix combining the fuzzified similarity measurements Fuzzy rules are activated at different degree of truth depending on the membership of the similarity measurements to fuzzy sets 17
- 18. Fuzzy Logic (Defuzzification) How to get an human friendly similarity score combining the similarity ratings measurements ? Defuzzify the fuzzy rules sets activated Using the « center of gravity » method 18
- 19. Visualisation 19
- 20. Visualisation of spatio-temporal data How to display spatio-temporal patterns and qualified positions/trajectories ? 3D space/time cube ? 20
- 22. Visualisation (2D analysis) 22
- 23. Conclusion Model of trajectory, itineraries and matching tools General methodology Data mining : spatio-temporal patterns Position and trajectory classification using fuzzy logic 23
- 24. Future work Improve statistics analysis (skewness/kurtosis) Detect multimodal groups of trajectories Investigate patterns generalization (aggregation ?) Consider more similarity measurements (heading, speed) Extend to trajectories partial matching, data streams, real time analysis Improve geovisualisation of outliers ... 24
- 25. Questions ? L. Etienne, T. Devogele, A. Bouju. In Shi, Goodchild, Lees & Leung (Eds.) Advances in Geo-Spatial Information Science, Chap. Modeling Space and Time, Spatio-temporal Trajectory Analysis of Mobile Objects Following the same Itinerary. CRC Press, Taylor & Francis Group, ISPRS Orange book series, ISBN 978-0-415-62093-2, pages 47-58, 2012. 25
- 26. Plateform programming PostgreSQL / PostGIS database Model & data integration (60 Gb of raw AIS data frames from different sources, 6 month ) PostGIS spatial functions & indexes PL/PgSQL, PL/C, PL/Java programming Java Spatio-temporal pattern extraction & similarity measurements Fuzzy logic Statistics Matlab Web PHP/HTML/JS/AJAX (Ajax Push Engine) GeoServer WFS/WMS Openlayers KML 26
- 27. Related publications Book chapters T. Devogele, L. Etienne, C. Ray, and C. Claramunt. In C. Renso, S. Spaccapietra & E. Zimányi (Eds.) Mobility Data: Modeling, Management, and Understanding, Chapter Mobility Applications, Maritime Applications. Cambridge press, to be published in 2012. L. Etienne, T. Devogele, A. Bouju. In Shi, Goodchild, Lees & Leung (Eds.) Advances in Geo-Spatial Information Science, Chap. Modeling Space and Time, Spatio-temporal Trajectory Analysis of Mobile Objects Following the same Itinerary CRC Press, Taylor & Francis Group, ISPRS Orange book series, ISBN 978-0-415-62093-2, pages 47-58, 2012. 27
- 28. Related publications International conferences L. Etienne, C. Ray, and G. Mcardle. Spatio-temporal visualisation of outliers. Proceedings of the international workshop on Maritime Anomaly Detection (MAD), pages 119–120, 2011. L. Etienne, T. Devogele, and A. Bouju. Spatio-temporal trajectory analysis of mobile objects following the same itinerary. Proceedings of the International Symposium on Spatial Data Handling (SDH), pages 86–91, 2010. A. Lecuyer, J.M. Burkhardt, and L. Etienne. Feeling bumps and holes without a haptic interface: the perception of pseudo-haptic textures. Proceedings of the SIGCHI conference on Human factors in computing systems, pages 239–246, 2004. 28
- 29. Europe map 29
- 30. Passenger ships 30
- 31. Calais - Dovers 31
- 32. Dover straits 32