Research

Topics

I am interested in Spatial Databases, with a particular focus on Spatial Crowdsourcing, Ridesharing, Spatial Advertising. Recently, I also work on Graph Databases, Blockchain and Knowledge Graph.

Code

I try to publish code promptly. You are welcome to contribute to the projects and cooperate with me and my team.

• VSAG: VSAG is a vector indexing library used for similarity search. The indexing algorithm allows users to search through various sizes of vector sets, especially those that cannot fit in memory. The library also provides methods for generating parameters based on vector dimensions and data scale, allowing developers to use it without understanding the algorithm’s principles. VSAG is written in C++ and provides a Python wrapper package called pyvsag. We publish a VLDB 2025 paper in industrial track to introduce the VSAG library: VSAG: An Optimized Search Framework for Graph-based Approximate Nearest Neighbor Search.
• Latex Template for CS papers: a simple latex template of cs papers for junior PG students. Many other useful materials are also put in the repository.
• gMission: a general spatial crowdsourcing platform.
• Spatial Crowdsourcing Dataset Generator: a toolbox to generate the synthetic datasets for spatial crowdsourcing applications.
• Benchmark for Task Assignment in Spatial Crowdsourcing: a benchmark to test the representitive algorithms for task assignment in spatial crowdsourcing. Details in our VLDB 2018 experiment analyses paper: An Experimental Evaluation of Task Assignment in Spatial Crowdsourcing.
• Queueing-Theoretic Framework for online Car-hailing: a framework to maximize the served requests of online car-hailing platforms using a queueing theoretic framework with DNN spatiotemporal prediction models. Details in our ICDE 2019 short paper: A Queueing-Theoretic Framework for Vehicle Dispatching in Dynamic Car-Hailing. An updated and enhanced framework for our VLDB 2021 full paper can be found here!
• Demand-Aware Route Planning for Shared Mobility Services: a framework to maximize the overall uitility of online ridesharing platforms using a demand-aware insertion route planning basic operator. Details in our VLDB 2020 paper: Demand-Aware Route Planning for Shared Mobility Services.
• GridTuner: a out-of-box tool to tune the size of grids for spatiotemporal prediction models to maximize the utility of the applications. Details in our ICDE 2022 paper: GridTuner: Reinvestigate Grid Size Selection for Spatiotemporal Prediction Models.
• Efficient K-clique Listing: source code of our ICDE 2022 paper: Efficient k-clique Listing with Set Intersection Speedup.
• Efficient Non-Learning Similar Substrajectory Search: source code of our VLDB 2023 paper: Efficient Non-Learning Similar Subtrajectory Search, which proposes a novel algorithm to search for the exact most-similar subtrajectory from a set of data trajectory for the given query trajectory under most widely used distance functions.

See the licensing terms within each project’s codebase for the requisite legal details.