Research Overview

Multi-modal data, such as trajectories, graphs, vectors, and text, is becoming increasingly common in our daily lives around the cities. To make different kinds of data usable, various databases are currently needed at a high cost, including relational databases, spatial databases, graph databases, etc. To reduce the cost on database communication and also save money on buying multiple databases, we should manage those multi-modal data in a single smart database system.

My current research is to build a multi-modal database management system for cross-disciplinary applications, such as self-driving, public transport, and tourism planning. Inevitably, such a system will be more and more complex in terms of pipeline configuration, including data cleaning, parameter setting, choice of similarity measures, etc. To make the system easy to use by various customer groups, I also want to achieve auto-configurations of learning and analytic pipelines, including data discovery, search, clustering, and routing.

Over the last several years, I also spent most of my time on trajectory data management, and an overview of trajectory data research can be found below (click to access our latest survey).

Wang, Sheng, Zhifeng Bao, J. Shane Culpepper, and Gao Cong. "A survey on trajectory data management, analytics, and learning." ACM Computing Surveys (CSUR) 54, no. 2 (2021): 1-36.