The tool we built for PingCAP’s TiDB Optimizer in automatically detecting the reasons behind query regression is accepted by VLDB22 demo track. The tool has been adopted within TiDB and signficantly saves optimizer developers’ time in analyzing the plan regression.
Our work titled “Representative Routes Discovery From Massive Trajectories” is accepted by KDD, Applied Data Science Track.
Our system work “VRE: A Versatile, Robust, and Economical Trajectory Data System” is accepted by VLDB. This is a joint work with Alibaba and the system core part has been deployed into GANOS, Alibaba’s spatial-temporal database engine. To our best knowledge, this is the first system that supports all the five complex queries and five widely-used distance metrics on large-scale trajectory data of various profiles (spatial span, point density, trajectory density), generated from the domains of aircraft, seafaring, and vehicles. Kudos to my student Hai!
Invited to serve the Associate Editor of ACM Transactions on Spatial Algorithms and Systems 2022-2025.
Our work “Equitable Public Bus Network Optimization for Social Good: A Case Study of Singapore” is accepted by ACM FAccT, a flagship venue for fairness and accountability of socio-technical systems. We introduce a comprehensive dataset based on Singapore consisting of the road network data, public transport data, and social demographics data. We propose a range of efficiency metrics and equity metrics and perform an in-depth look of the dataset to show the intricacies of the Singapore bus network, one of the best-in-class public transport systems worldwide.
Our work “Local Clustering over Labeled Graphs: An Index-Free Approach” is accepted by IEEE ICDE 2022.
Our work “Approximate Range Thresholding” is accepted by SIGMOD 2022.
Outstanding Senior PC Award, by ACM WSDM 2022.
An ARC Discovery project is funded and four PhD scholarships are open until filled. Another two scholarships are available for a project on large-scale transaction processing.
Our work “Dynamic Ridesharing with Minimal Regret: Towards an Enhanced Engagement Among Three Stakeholders” will appear in TKDE.
I will serve the Associate Editor of PVLDB Vol 16 and that of SIGMOD 2023, as well as the Senior PC of KDD 2022.
Our scalable data science work “Points-of-Interest Relationship Inference with Spatial-enriched Graph Neural Networks” is accepted by PVLDB 2022.
Congratulations to my student Sheng in joining Wuhan University as an Assoc Professor.
VLDB 2021 is happening! We organize 17 round tables, covering a wide range of trending topics such as Systems for ML, Network Embedding and Data Preparation for ML. Kudos to everyone that helps make it happen!]
Our work titled “Robust Road Network Representation Learning: When Traffic Patterns Meet Traveling Semantics” is accepted by ACM CIKM 2021.
Our work on building a Framework to support continuous range queries over Multi-Attribute Trajectories is accepted by IEEE TKDE.
Invited to be the RoundTable Chair of VLDB 2021!
Our work on how Context or Knowledge can benefit healthcare Question Answering is accepted by IEEE TKDE.
Our work on minimizing regret for outdoor advertising is accepted by SIGMOD 2021. Congratulations to my PhD student Yipeng Zhang.
I am invited to be the Program Committee Co-chair of ACM CIKM 2021 (Demo Track).
Our survey on Advancing the DBMS Query Optimizer is available. This survey focuses on Cardinality Estimation, Cost Model, and Plan Enumeration.
Our work about triangle counting over graph data is accepted by TKDE.
Our paper on transit planning is directly accepted by SIGMOD 2021.
It is an honor to be invited as a Senior PC member of KDD 2021.
Our work “Towards an Efficient Weighted Random Walk Domination” is accepted by PVLDB’21.
Our work “AOI-shapes: an efficient footprint algorithm to support visualization of user-defined urban areas of interest” is accepted by ACM TiiS. This work is a typical effort drawing inspirations between visualization and big data algorithm.
Our work “Towards an Optimal Bus Frequency Scheduling: When the Waiting Time Matters” is accepted by TKDE.
Our work “Towards Efficient Motif-based Graph Partitioning: An Adaptive Sampling Approach” is accepted by ICDE 2021. This is the first attempt to solve this problem without employing exact edge weight computations.
Our work “On the Efficiency of K-Means Clustering: Evaluation, Optimization, and Algorithm Selection” is accepted by PVLDB 2021.
How to effectivenly recommend indexes Using Deep Reinforcement Learning? Check our CIKM’20 work and the code.
Our system prototype on bus (re)scheduling assistant, FASTS, is online. Check our introductory video and details are available at our VLDB’20 demo paper “FASTS: A Satisfaction-Boosting Bus Scheduling Assistant”.
Our Paper on traffic prediction is accepted by ICDE 2021.
Our Paper “Incremental Preference Adjustment: a Graph Theoretical Approach” is accepted by VLDB Journal. Our proposed techniques can be applied to any vector embedding based preference learning model and incrementally adjust the preference for every single user via a few interactions. A very interesting case study has also been conducted. Check our spotlight talk.
It is an honor to serve the Registration Chair and the PC member of SIGMOD 2021.
“A Survey on Modern Deep Neural Network for Traffic Prediction: Trends, Methods and Challenges” is accepted by TKDE. Check an Early Access Version.
Our paper “Spatial Object Recommendation with Hints: When Spatial Granularity Matters” got accepted by SIGIR’20. This is a joint work with Baidu Research using real-world Baidu Maps user log data.
Our Paper “Effective Travel Time Estimation: When Historical Trajectories over Road Networks Matter” has been accepted by SIGMOD 2020.
We have one paper “Crowdsourced Collective Entity Resolution with Relational Match Propagation” accepted by IEEE ICDE 2020.
I am promoted to Associate Professor, effective from Jan 1, 2020. Thanks for the recognition from the peers and the university promotion panel!
We get an ARC DP20 grant to support our research on building a next-generation map service for an interactive exploration of Massive Geo-spatial data. Thanks ARC!
We have two papers “Online Anomalous Trajectory Detection with Deep Generative Sequence Modeling” and “Temporal Network Representation Learning via Historical Neighborhoods Aggregation” accpeted by IEEE ICDE 2020
Our paper “Optimizing Impression Counts for Outdoor Advertising” got the Best Paper Award Runner Up in ACM SIGKDD 2019
Our paper “Fast Large-Scale Trajectory Clustering” is accepted by PVLDB 2020
We are grateful to a new Google Research Award. Thanks Google!
Our paper ‘Trajectory-driven Influential Billboard Placement’ is selected as one of the Best Papers of SIGKDD