Tourist Behavior-based Early Warning Scheme for Wisdom Tourisms

Jinghe Zhang, Qianwei Chen, Xiaowei Wu, Yue Hu, Weidong FangAdvances 2023 · DOI · OpenAlex

TL;DR

This paper proposes a location-aware early warning system for "Wisdom Tourism" that uses mobile GPS data and predictive algorithms to nudge tourists away from dangerous or uncivilized behaviors. By combining Newton interpolation for path prediction with interactive virtual reality (VR) elements, the scheme aims to replace static warning signs with real-time, personalized safety interventions.

What Problem Does This Paper Address?

As global tourism scales, scenic spots face a dual challenge: ensuring visitor safety in dangerous terrain and curbing "uncivilized" behaviors (like littering or smoking in fire-prone areas). Traditional methods—static signs and physical staff—often fail during peak seasons due to "signage fatigue" or sheer crowd volume.

The authors argue that current "Wisdom Tourism" (the integration of IT into travel) focuses too much on digitization and not enough on the proactive management of visitor behavior. There is a need for a system that doesn't just track where a tourist is, but predicts where they are going and intervenes before a mishap occurs.

Methods

The proposed scheme relies on a multi-layered approach to location tracking and behavioral intervention:

  1. Position Prediction via Newton Interpolation: Rather than relying on a simple "ping" of current coordinates, the system uses Newton interpolation to estimate a user’s position at time $t$ based on their three previous historical locations ($t-1, t-2, t-3$).
  2. Trend Prediction & Key Weighted Areas: The algorithm defines a "Circle C" around the user's current location, representing the maximum distance they could travel in a given interval. Within this circle, it identifies a "key weighted area"—a wedge-shaped zone based on the user's current trajectory.
  3. Monte Carlo Sampling: The system uses Monte Carlo methods to sample points within the predicted path. Points falling within the "key weighted area" are given higher weights, allowing the system to trigger warnings more accurately if the predicted path intersects with a dangerous zone.
  4. Virtual Alarm Operations: When a tourist enters (or is predicted to enter) a warning zone, the app triggers a "virtual alarm." This ranges from simple video playback of correct behaviors to interactive 3D scenes. For example, a user might be shown a VR simulation where they must "click" a water bottle to extinguish a discarded cigarette butt, earning "tourist points" as a reward for the correct action.
  5. Time-Based Fallbacks: For users with disabled GPS or poor signal, the system includes a time-variable trigger that periodically serves safety content to ensure no visitor is left without guidance.

Datasets

Not discussed in the paper. The authors focus on the mathematical framework and the system architecture rather than training on a specific historical dataset or benchmarking against existing tourism datasets.

Key Insights and Ideas

  • Predictive vs. Reactive: The core innovation is the shift from geofencing (reacting when someone crosses a line) to path prediction. By using Newton interpolation, the system attempts to "see" the tourist's intent.
  • Gamified Safety: Instead of a jarring siren or a text alert, the system uses interactive VR modules. The "aha moment" here is the use of a reward system (points) to incentivize tourists to engage with safety training, turning a boring warning into a mini-game.
  • The "Key Weighted Area": By combining the maximum possible speed of a tourist with their current heading, the authors create a probabilistic map of where the user is most likely to be, which helps reduce "false alarm" warnings for tourists who are near a danger zone but moving away from it.

Key Results

The paper is primarily a theoretical and architectural proposal. Consequently, it lacks quantitative experimental results (such as "X% reduction in accidents"). However, it provides:

  • A complete logic flow for a mobile-based early warning system.
  • Mathematical proofs for position estimation and trajectory weighting.
  • Qualitative Framework: A structured method for integrating "Wisdom Tourism" infrastructure with police and park management systems to improve overall work efficiency.

Limitations and Future Work

  • Hardware & Battery Constraints: The paper assumes tourists will keep a GPS-heavy app running. In remote scenic areas, GPS accuracy can be poor, and high-frequency location polling significantly drains mobile battery life.
  • User Adoption: The effectiveness of the scheme depends on tourists actually downloading and using the app. The authors suggest "points" as an incentive, but the friction of app installation remains a hurdle.
  • Empirical Validation: The paper does not include a field study or a pilot program. Future work would benefit from testing this algorithm in a real-world park to see how tourists actually respond to "virtual alarms" versus traditional signage.

Summary updated

Abstract

In this paper, we propose a warning scheme for behavior early warning in scenic spots.In the scheme, it determines whether to send an alarm reminder based on the current location of the visitor's mobile terminal.The warning scheme includes obtaining the warning location information based on preset dangerous location data, as well as obtaining the current location information of the visitor at a preset time interval, and setting the current location of the visitor.The location information is compared with the warning location information, and if the comparison results are consistent, a virtual alarm operation is triggered.By comparing the current location of the tourist with the warning location information, it is determined whether the virtual warning operation is triggered, so that tourists can obtain real-time warning information in the scenic spot, for reducing the probability of uncivilized or dangerous behaviors in the scenic spot, improving the warning operation, and making sure that it is safe and civilized to travel in the end.
BibTeX
@inproceedings{zhang2023tourist,
  author = {Jinghe Zhang and Qianwei Chen and Xiaowei Wu and Yue Hu and Weidong Fang},
  title = {{Tourist Behavior-based Early Warning Scheme for Wisdom Tourisms}},
  booktitle = {Advances},
  year = {2023},
  doi = {10.2991/978-94-6463-058-9\_3},
}

Loading related content...

Figure