Pattern Recognition for Spatio-temporal Dynamics

Description

The widespread use of location-tracking technologies such as GPS and GSM networks has made it easy for individuals to create a spatial-temporal record of their whereabouts. This has created an opportunity to extract valuable insights from this type of data. This project aims to understand and capture individual’s general life styles and regularity. Mining individual life patterns has the potential to be applied in a wide range of scenarios. One important application of life pattern analysis is in location-based recommendation systems. By understanding an individual’s life pattern, recommendation engines can provide more personalized recommendations for products and services that are relevant to their lifestyle. We use large-scale GPS location data and spatial temporal point processes to uncover the hidden spatiotemporal patterns from the locations.


People


Haowen Lin

CS PhD Student, USC