.png){kind=icon}
返回
Relaxed group pattern detection over massive-scale trajectories
.png){kind=icon}
PRE
AI
DOI：10.1016/j.future.2023.02.028.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The problem of detecting co-movement patterns from trajectories has been extensively studied by data science community. Existing methods are based on filter-and-refine framework, indexing, location point clustering, and search algorithms. Although these techniques are widely used, they require huge computation efforts to find patterns with complex group definition, making them incapable of producing timely results over massive-scale trajectories. In real-life scenarios, especially in pandemic alert and prevention, it is important to produce real-time and high-quality results. In this light, we simplify and relax the definition of co-movement pattern to support fast detection without loss of generality. Specifically, we define co-movement pattern through companion group that consists of at least m moving objects traveling together for k time periods. If an object can form multiple companion groups with others, it belongs to the companion group with the maximum member size. Given a collection of moving object trajectories, the exact algorithm finds companion groups by combining qualified co-movement pairs, which is time-consuming. To enable efficient discovery, a fast community-aware approximate algorithm is developed. Extensive experiments on two real-world datasets are conducted to verify the effectiveness and efficiency of our proposal.(c) 2023 Elsevier B.V. All rights reserved.
Keyword:
Co-movement pattern mining
Trajectory clustering
Group pattern detection
期刊
F
IF:
6.1
论文数:
6.9K
被引数:
2.3W
机构
暂无机构信息
引用论文
How to find appropriate automobile exhibition halls: Towards a personalized recommendation service for auto show.png)
NEUROCOMPUTING
NEUROCOMPUTINGIF6.5
PREAI
分享收藏A Novel Framework for Online Amnesic Trajectory Compression in Resource-Constrained Environments
OAAI
OAAI分享收藏
分享收藏分享收藏
查看更多内容