Tourism Route Planning Based on Data Analysis and Mathematical Modeling

Chang Liu, Rongjie Cai, Z. J. Xiao, Guanliang Chen, Xin Guo, Yingzhi Ma, Xiuyuan YangMathematical 2025 · DOI

This paper, published in Mathematical Modeling and Algorithm Application, tackles a modern logistical challenge: how can a foreign tourist make the most of China’s 144-hour visa-free transit policy? By combining data analysis with greedy optimization, the authors propose a framework for planning the "perfect" six-day trip.

TL;DR

The paper develops a mathematical model to optimize multi-city tourism routes for foreign visitors under China's 144-hour visa-free policy. Using a greedy algorithm, the authors balance scenic spot ratings against travel time and costs to generate an efficient, high-satisfaction itinerary starting and ending in Guangzhou.

What Problem Does This Paper Address?

With the rise of the "city is not city" social media trend and the expansion of visa-free transit, China has seen a surge in international "layover" tourism. However, planning a trip within a strict 144-hour (6-day) window is a complex optimization problem.

Tourists face a classic trade-off: they want to visit the highest-rated attractions (maximizing "utility") while minimizing the time and money spent on high-speed rail (HSR) travel. The authors aim to provide a scientific basis for route planning that moves beyond simple intuition, accounting for constraints like daily visit limits, HSR speeds, and fixed budgets.

Methods

The researchers framed the problem as a constrained optimization task.

  • The Objective Function: The model seeks to maximize a comprehensive evaluation score ($S$). This score is a weighted function of three primary factors: the quality of scenic spots ($R_i$), the travel time between cities ($D_{ij}$), and the transportation costs ($F_{ij}$).
  • Constraints: The model enforces a hard limit of 144 total hours. It also assumes a realistic "tourist workday" of 8 hours per day and limits visitors to one top-tier scenic spot per city to keep the itinerary manageable.
  • The Algorithm: To solve the model, the authors employed a Greedy Algorithm. Instead of calculating every possible permutation of cities (which is computationally expensive), the algorithm makes the locally optimal choice at each stage—picking the next city that offers the best balance of high scores and low travel costs from the current location—until the 144-hour limit is reached.

Datasets

The paper does not cite a specific public dataset (like a Kaggle set), but rather utilizes a curated set of parameters for Chinese cities and scenic spots. These include:

  • Fixed scenic spot ratings and ticket prices.
  • Suggested "play times" for specific attractions.
  • A standardized high-speed rail matrix (assuming an average speed of 300 km/h and a cost of 100 yuan per hour of travel).

Key Insights and Ideas

  • HSR as the Backbone: The model assumes high-speed rail is the sole mode of intercity transport. While a simplification, this reflects the reality of China’s infrastructure, where HSR is often more efficient than flying for regional hops.
  • The "8-Hour Rule": By capping daily tourism at 8 hours, the model avoids "burnout" itineraries that are mathematically optimal but physically impossible for a human to enjoy.
  • Efficiency vs. Global Optima: The choice of a greedy algorithm suggests that for tourism planning, a "good enough" fast solution is often preferable to a "perfect" solution that takes hours to compute, especially when dealing with the dynamic nature of travel.

Key Results

The authors successfully generated a loop-style itinerary starting and ending in Guangzhou.

  • Route Visualization: The resulting maps (Figures 1 and 2 in the paper) show a logical progression through neighboring high-value cities, ensuring the tourist returns to their port of entry within the 144-hour window.
  • Balanced Optimization: The "optimized" version of the route (Section 4.2) demonstrates that by adjusting weights ($w_1, w_2, w_3$), the model can prioritize budget-friendly travel without significantly sacrificing the quality of the attractions visited.

Limitations and Future Work

The authors are transparent about several "real-world" gaps in their current model:

  • The "Last Mile" Problem: The model ignores the time it takes to get from a high-speed railway station to the actual scenic spot, which can be significant in mega-cities.
  • Single-Mode Transport: It does not account for flights or ride-sharing, which might be faster or cheaper for certain routes.
  • Subjectivity: The weights for "quality" vs. "cost" are currently set by the researchers. Future iterations could use machine learning to personalize these weights based on individual user preferences (e.g., a "history buff" vs. a "nature lover").
  • Static Variables: The model doesn't account for seasonal closures, weather, or fluctuating ticket prices. Integrating real-time data APIs would be the logical next step for a production-ready version of this tool.

Summary updated

Abstract

With the practical implementation of the visa-free transit policy in recent years, more and more foreign tourists are traveling in China. Through data analysis and mathematical modeling, this paper solves a series of problems that foreign tourists may face in the process of traveling in China, and provides detailed tour routes and statistical results. These research results not only provide practical reference for foreign tourists, but also provide valuable data support and decision-making basis for the development of tourism in China.
BibTeX
@inproceedings{liu2025tourism,
  author = {Chang Liu and Rongjie Cai and Z. J. Xiao and Guanliang Chen and Xin Guo and Yingzhi Ma and Xiuyuan Yang},
  title = {{Tourism Route Planning Based on Data Analysis and Mathematical Modeling}},
  booktitle = {Mathematical},
  year = {2025},
  doi = {10.54097/8jqrqs34},
}
Figure