Two-Stage Framework for Inferring Air Trips from Incomplete LBS Data

This framework recovers domestic US air trips from mobile location data by combining heuristic rules for high-confidence sightings with a probabilistic model for latent routes. The model integrates spatial proximity, temporal compatibility, and flight schedules to validate results against T-100 Domestic Segment benchmarks.