Abstract:
Understanding point-of-interest (POI) mobility patterns improves mobile business intelligence. While many studies have attempted this, most use POI check-ins to mine mobility patterns, which is ineffective due to data sparsity.
This project directly annotates POIs from raw user-generated mobility records to improve POI-based human mobility for mining. Neural context fusion integrates context factors in POI-visiting behaviors. Representation learning evaluates preference and transition factors.
We use an attention mechanism for randomized raw mobility transitions. Our data-driven approach continues to use domain knowledge factors like distance, time, and popularity. Feed-forward neural networks automatically fuse factors.
A multi-head architecture improves model expressiveness. Our experimental study on two real-life data sets shows that our approach outperforms state-of-the-art baselines by at least 32% in accuracy. A POI recommendation example shows the utility of POI-based human mobility.
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
