Mining and Simulating of Tourists'Behavior Patterns in Multi-scale Red Tourism along the Long March Route
Tourists'spatial behavior pattern is a key theoretical issue in tourism management.Amid China's efforts to construct the Long March National Cultural Park,scientifically integrating red tourism resources,mastering the situation of regional integration and understanding the red tourism behavior pattern holds the key to high-quality development of red tourism along the Long March route.The exploration of tourists'behavior patterns in multi-scale red tourism can enrich the theoretical research in geography on human leisure activity spaces and decision-making behaviors and offer support for meeting the practical needs of spatial planning and tourism route optimization in building the Long March National Cultural Park.Therefore,there is a pressing need for the academic community to conduct pertinent studies in light of the current scenario.The spatiotemporal big data generated by tourists brings new opportunities for research on their spatiotemporal behaviors.Compared with traditional data collection methods,spatial and temporal data utilizing various positioning technologies can be acquired in a shorter time,in larger quantities and with higher precision.It is worth noting that the analysis methods of tourism spatiotemporal behaviors developed based on traditional data fail to meet the demand for big data analysis.Although interdisciplinary research methods such as deep learning,statistical physics,and complex networks provide brand-new research tools for the study of tourists'behaviors,the adoption of cutting-edge technologies in tourism research still needs further exploration and expansion.Developing big data analysis methods applicable to tourism scenarios that incorporate the characteristics of tourism big data,reviewing existing scientific questions,and asking and answering new ones are gradually becoming an important part of tourist behavior study.To create a dataset of red tourist scenic spots and trajectories along the Long March route,this study acquired 1 576 093 tourism GPS trajectories from 2012 to 2022,extracted stay points,and selected red tourism trajectories.Subsequently,this study established a trajectory data mining framework applicable to the tourism scenarios,which uses the HDBSCAN model to mine the behavior patterns of red tourists from large,medium and small spatial scales respectively,and classifies red tourism patterns taking types of scenic areas into account.To determine the resilience of the mined red tourism patterns,a simulation study of the pattern network under abnormal circumstances was performed.The main conclusions are:(1)The tourist trajectory analysis method and framework constructed in this study are feasible.The larger the scale of red tourism activities along the Long March,the fewer the patterns.There are respectively 17,56 and 81 red tourism patterns at large,medium and small scales;(2)Eight main interconnected areas have been formed in the areas along the Long March route,which features coordination between two to four provinces at a large scale,interconnection between three to six cities at a medium scale,and linkage between one to three Long March-related scenic spots at a small scale.(3)From the perspective of pattern network resilience,the medium-scale red tourism patterns are the most resilient,followed by the small-scale and large-scale ones,with a greater resilience difference between the medium-scale and small-scale patterns.The simulation model can accurately portray the scenarios of tourism destinations perturbed by abnormal situations,and the constructed resilience measurement model effectively measures the resilience of the pattern network,which is a useful exploration for spatial behavior pattern assessment.
spatio-temporal behavior patternspattern network resiliencered tourismthe Long Marchunsupervised learning
NETL
NSTL
万方数据
