Abstract:Given a set of user-specified locations and a massive trajectory dataset, the task of mining spatio-temporal reachable regions aims at finding which road segments are reachable from these locations within a given temporal period based on the historical trajectories. Determining such spatio-temporal reachable regions with high accuracy is vital for many urban applications, such as location-based recommendations and advertising. Traditional approaches to answering such queries essentially perform a distance-based range query over the given road network, which does not consider dynamic travel time at different time of day. By contrast, we propose a data-driven approach to formulate the problem as mining actual reachable regions based on a real historical trajectory dataset. Efficient algorithms for the Single-location spatio-temporal reachability Query (S-Query) and the Union-of-multi-location spatio-temporal reachability Query (U-Query) were presented in our recent work. In this paper, we extend the previous ideas by introducing a new type of reachability query with multiple sources, namely, the Intersection-of-multi-location spatio-temporal reachability Query (I-Query). As we demonstrate, answering I-Queries efficiently is generally more computationally challenging than answering either S-Queries or U-Queries because I-Queries involve complicated intersect conditions. We propose two new algorithms called the Intersection-of-Multi-location Query Maximum Bounding region search (I-MQMB) algorithm and the I-Query Trace Back Search (I-TBS) algorithm to efficiently answer I-Queries, which utilize an indexing schema composed of a spatio-temporal index and a connection index. We evaluate our system extensively by using a large-scale real taxi trajectory dataset that records taxi rides in Shenzhen, China. Our results demonstrate that the proposed approach reduces the running time of I-Queries by 50 percent on average compared to the baseline method.

Mining Spatio-Temporal Reachable Regions with Multiple Sources over Massive Trajectory Data

Search Model of the Region with the Maximum Coverage Value Based on Trajectory Data

Partition-based range query for uncertain trajectories in road networks

Time-Based Trajectory Data Partitioning for Efficient Range Query.

The Discovery of Personally Semantic Places Based on Trajectory Data Mining

Mining interesting locations and travel sequences from GPS trajectories

Mining Trip Attractive Areas Using Large-Scale Taxi Trajectory Data

SemanticTraj: A New Approach to Interacting with Massive Taxi Trajectories.

Towards Efficient Search for Activity Trajectories

User Oriented Trajectory Search for Trip Recommendation

A Hilbert-GeoSOT Spatio-Temporal Meshing and Coding Method for Efficient Spatio-Temporal Range Query on Massive Trajectory Data

A visual reasoning approach for data-driven transport assessment on urban roads

Personalized trajectory matching in spatial networks

Dynamic Spatio-temporal Access Queries using Semi-Supervised Regression

Road-Aware Indexing for Trajectory Range Queries

Non-Uniform Spatial Partitions and Optimized Trajectory Segments for Storage and Indexing of Massive GPS Trajectory Data

Efficient Path Query Processing over Massive Trajectories on the Cloud

Browsing target extraction and spatiotemporal preference mining from the complex virtual trajectories

Efficient Mining of Hotspot Regional Patterns with Multi-Semantic Trajectories

Optimizing segmented trajectory data storage with HBase for improved spatio-temporal query efficiency

Querying Massive Trajectories by Path on the Cloud