An Improved HMM-Based Approach for Planning Individual Routes Using Crowd Sourcing Spatiotemporal Data

Area of Science:

• Computer Science
• Geographic Information Science

Background:

• Location-based services (LBS) are rapidly developing, increasing interest in personalized travel routes.
• Existing LBS route planning lacks dynamic scalability and fails to investigate environmental constraints versus personal choices.

Purpose of the Study:

• To propose an improved Hidden Markov Model (HMM)-based method for personalized route planning.
• To integrate dynamic public preferences, individual interests, and road network space within a spatiotemporal framework.
• To develop a dual-layer mapping structure for connecting preferences to Points of Interest (POIs) in realistic road networks.

Main Methods:

• Utilizing crowdsourced spatiotemporal data.
• Implementing an improved Hidden Markov Model (HMM).
• Developing a novel dual-layer mapping structure.

Main Results:

• The proposed method enables flexible route planning across diverse spatiotemporal contexts.
• The approach generates routes that align with group perceptions and natural selection.
• A case study in Changsha city validated the method's effectiveness.

Conclusions:

• The improved HMM method offers a scalable and personalized approach to route planning.
• Integrating public and individual preferences enhances route relevance and practicality.
• The dual-layer mapping structure effectively bridges user preferences with real-world POIs.