Mining user similarity based on location history
Proceedings of the 16th ACM SIGSPATIAL international conference on Advances in geographic information systems
Trajectory Outlier Detection: A Partition-and-Detect Framework
ICDE '08 Proceedings of the 2008 IEEE 24th International Conference on Data Engineering
Efficient anomaly monitoring over moving object trajectory streams
Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining
Finding similar users using category-based location history
Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems
| Hi-index | 0.00 |
Human activity behaviors in urban areas mostly occur in interior places, such as department stores, office buildings, and museums. Understanding and characterizing human spatio-temporal interactive behaviors in these indoor areas can help us evaluate the efficiency of social contacts, monitor the frequently asymptomatic diseases transmissions, and design better internal structures of buildings. In this paper, we propose a new temporal quantity: 'Participation Activity Potential' (PPA) to feature the critical roles of individuals in the populations instead of their degrees in the corresponding complex networks. Especially for the people with high degrees (hubs in the network), Participation Activity Potential which is directly from the statistics of their daily interactions, can easily feature the rank of their degree centrality and achieve as high as 100% accuracy rating without building the corresponding networks by high-complexity algorithms. The effectiveness and efficiency of our new defined quantity is validated in all three empirical data sets collected from a Chinese university campus by the WiFi technology, a small conference and an exhibitions by the RFID technology.