Abstract:
Location-based data mining has been extensively studied as spatial-temporal trajectory data accumulates. This field studies self-supervised pre-training to learn location embedding vectors. Pre-trained embedding vectors can use abundant unlabeled trajectory data to benefit downstream tasks.
Most methods ignore temporal information in trajectories’ visited times. Since human activities are highly regulated by specific times of day, temporal information can reflect some intrinsic characteristics of locations, so location embedding vectors must fuse them.
We propose a CBOW-based Time-Aware Location Embedding (TALE) pre-training method that incorporates temporal information into location embedding vectors. Hierarchical Softmax calculates temporal information using a novel temporal tree structure.
We apply the learned embedding vectors to three downstream location-based prediction tasks—location classification, location visitor flow prediction, and user next location prediction—to test TALE. Our TALE model improves downstream task performance in four real-world user trajectory datasets.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
