Using interactions and dynamics for mining groups of moving objects from trajectory data

Abstract

Recent advances in tracking technology enable the gathering of spatio-temporal data in the form of trajectories. Analyzing trajectories can convey knowledge useful for prominent applications and designing computational solutions for mining groups of moving objects may turn out to be a valuable means for a wide class of problems related to mobility. The task of group mining has been investigated by considering mostly the spatial closeness and similarity of the trajectories, while little attention has been paid to the relationships between the trajectories and time-changing nature of the trajectories. The relationships may provide evidence of interactions between the moving objects. The time-changing nature may provide evidence of dynamics of the movements. Therefore, interactions and dynamics can be sources of information that one can consider to discover new forms of groups. In fact, groups of objects may be of interest not only when the objects move together or move close from each other, but also when they come from different places, change direction, join together and then move away from each other. Motivated by this, we introduce the concept of crews and propose a computational solution to discover crews. A crew gathers moving objects with similar interactions and similar dynamics. The proposed computational solution relies on i) new movement parameters, which explicitly consider interactions and dynamics, and ii) a distance-free clustering algorithm, which groups objects based on the similarity of the movement parameters. We conduct extensive experiments on real-world trajectory data, present a quantitative evaluation of the quality of the crews and perform comparisons with a baseline algorithm and with an algorithm of group pattern mining. The empirical results provide interesting insights on the relevance of some parameters in the construction of the crews.

An experimental evaluation of grouping definitions for moving entities

“…For example, the loose traveling companions [33] do not need to use the "within-group" connectivity for the entire duration of a group. Besides distance between entities, crews [29] also use many more aspects (e.g. speed, tortuosity, etc) to define the connectivity between entities.…”

“…Many different definitions have been suggested to model the collective movement of a "sufficiently large" set of entities that travel "together" for a "sufficiently long" period of time: flocks [4,13,44], mobile groups [16], moving clusters [21], moving micro-clusters [27], herds [15], convoys [19], swarms [28], gatherings [55], traveling companions [42], platoons [26], groups [5], refined groups [43], crews [29], and evolving companions [37].…”

Massively Parallel Discovery of Loosely Moving Congestion Patterns from Trajectory Data

“…Here, researchers noted that the interactions between moving objects would influence the mining of the group pattern as well as the dynamics. In a previous study [27], the concept of a crew was defined as a group of moving objects gathering with similar interactions and similar dynamics. The study also proposes a computational solution to discover crews from raw trajectory data.…”

Structured Medical Pathology Data Hiding Information Association Mining Algorithm Based on Optimized Convolutional Neural Network

“…where, x k,i ∈ X k is the set of position and velocity of data i on the horizontal and vertical axes, and i is the parent of data i. b k (h, i) is a compensation vector, which is the positional relationship between the data Q and its parent node; F and C are the data vector transfer matrix and the observation matrix, respectively; B is the data vector noise matrix; w and v represent the noise present in the CNN, respectively. The observed noise is subject to the positive distribution [39]- [41]. By analyzing the adjacency matrix, it is easy to confirm the association between the information in the medical information hiding in the pathological data group, such as the parent-child association [42]- [44].…”