Signature-Based Trajectory Similarity Join

Ta, Na; Li, Guoliang; Yong-qing, Xie; Li, Chang-Qi; Hao, Shuang; Feng, Jianhua

doi:10.1109/tkde.2017.2651821

Cited by 66 publications

(50 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, in Figure 3(b), Pps \ v1.p.s = {τ2, τ4}; thus, the margin of v1.p is 2. Moreover, a trajectory with a higher spatiotemporalsimilarity upper bound (Equation 19) is more likely to be the solution. So, ∀τ ∈ Pps \ q.s, the value of SimST(τ1, τ ).ub is proportional to the priority of query source q.…”

Section: Scheduling Strategymentioning

confidence: 99%

“…The TS-Join may bring significant benefits to a range of applications, including trajectory near-duplicate detection, data cleaning [2,19], ridesharing recommendation [16,17], friend recommendation [17], frequent trajectory based routing [13,19], and traffic congestion prediction. For example, a database may contain several copies of a trajectory or several similar trajectories.…”

Section: Introductionmentioning

confidence: 99%

“…Next, unlike existing trajectory similarity joins [2,3,6,10,19], the TS-Join is applied in spatial networks because in many practical scenarios, objects (e.g., commuters and vehicles) move in spatial networks rather than in Euclidean space. In spatial networks, only the network distance can reflect the real distance between two objects, and Euclidean distance may lead to errors.…”

Section: Introductionmentioning

confidence: 99%

“…(i) Different query spaces (Euclidean vs. network): these joins (e.g., [2,3,6,10,19]) are conducted in Euclidean space rather than in a spatial network. Thus, existing spatial indices (e.g., the R-tree [11]) and accompanying techniques are ineffective here.…”

Section: Introductionmentioning

confidence: 99%

“…(iii) Parallel processing: an experimental study [12] shows that existing centralized similarity join techniques (that do not take parallel processing into account) are far from efficient in processing very large data sets. Existing centralized trajectory similarity joins (e.g., [2,3,6,10,19]) can process at most 500 K trajectories (based on their experiments), while the TS-Join is able to process 10 M trajectories with a reasonable runtime (e.g., processing 10 M × 2 M trajectories for non-self join in 255 seconds and processing 10 M trajectories for self join in 540 seconds). A comparison between the TS-Join and existing studies is shown in Table I.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Trajectory similarity join in spatial networks

Chen²,

et al. 2017

View full text Add to dashboard Cite

The matching of similar pairs of objects, called similarity join, is fundamental functionality in data management. We consider the case of trajectory similarity join (TS-Join), where the objects are trajectories of vehicles moving in road networks. Thus, given two sets of trajectories and a threshold θ, the TS-Join returns all pairs of trajectories from the two sets with similarity above θ. This join targets applications such as trajectory near-duplicate detection, data cleaning, ridesharing recommendation, and traffic congestion prediction.With these applications in mind, we provide a purposeful definition of similarity. To enable efficient TS-Join processing on large sets of trajectories, we develop search space pruning techniques and take into account the parallel processing capabilities of modern processors. Specifically, we present a two-phase divideand-conquer algorithm. For each trajectory, the algorithm first finds similar trajectories. Then it merges the results to achieve a final result. The algorithm exploits an upper bound on the spatiotemporal similarity and a heuristic scheduling strategy for search space pruning. The algorithm's per-trajectory searches are independent of each other and can be performed in parallel, and the merging has constant cost. An empirical study with real data offers insight in the performance of the algorithm and demonstrates that is capable of outperforming a well-designed baseline algorithm by an order of magnitude.

show abstract

Section: Scheduling Strategymentioning

confidence: 99%