Key Research Highlights > Databases & Big Data Analytics > Management of Location-based Data Analytic and Real-Time Services
Kian Lee TAN
Management of Location-based Data Analytic and Real-Time Services
To analyse data generated by modern devices providing location technologies
We have met the objectives through the development of algorithms to mine and discover a variety of interesting patterns including: 1) convoy patterns to identify groups of objects moving together; 2) frequent route patterns to identify routes that are frequently traversed; and 3) meeting patterns to uncover regions where moving objects meet.
Additionally, we developed TsingNUS a real-time, location-based service system that has the following advantages over traditional location-aware applications: 1) location-aware search-as-you-type feature to enable answers to be continuously returned and refined as the query is typed letter-by-letter; 2) support for spatial-keyword queries for both AND and OR semantics; 3) incorporation of a continuous spatial keyword search that automatically refines the answers as users move around. Our algorithms and data structures resulted in superior real-time performance for very large datasets with over millions of point-of-interest.