以南京市为例,通过对旅游攻略的数据挖掘,构建景点间游客流动的关系矩阵,对流动格局、机制和模式进行了系统地分析,研究表明:1景点具有明显的层级性,市级中心景点在流动网络中处于核心地位;中山陵为扩散型景点,而新街口、夫子庙为集聚型景点,其他大部分景点为平衡型景点。2距离对景点间流动起着重要影响,流量距离衰减特征明显;市级中心景点流动涉及面较广,对整体流动具有较强的支配作用。3景点间流动是景点流入度、流出度和景点间距离共同作用的结果,其流动符合基于幂函数的威尔逊流动模型,且游客在城市内部景点间流动对距离的敏感度相对城市间、城市内部居民日常流动较低。4南京景点可划分为3个系统:中山陵景点系统、夫子庙景点系统和中心城区景点系统,并抽象出3种结构类型:单中心扩散结构、单中心集聚结构和多中心平衡结构;在流动模式上,主要流动多存在于主要节点之间,次要流动多存在于主要节点和次要节点之间以及次要节点之间,一般流动多存在于次要节点和一般节点之间以及一般节点之间。
Taking Nanjing as a case study, this paper calculates a matrix to describe the flows between scenic spots using data mining of tourism strategies and then systematically analyzes the structures, mechanisms and patterns of the flows. This study has drawn the following conclusions: (1) All the scenic spots form a hierarchy of several levels and the central ones are positioned as a core node on the flow network. Based on the quantity of inflows and outflows, Sun Yat-sen's Mausoleum is classified as a diffusion-type spot, Confucius Temple as a clustering-type one and Xinjiekou as a balance-type one. (2) The tourist flows have demonstrated a clear effect of distance decay. The scenic spots in inner city play a dominant role in influencing the overall flow. (3) The pattern of the flows is determined by the interactions between the intensities of inflows, outflows and the origin-destination distance. These flows follow the Wilson model represented as a power function. Compared with the inter-city and intra-city daily flows, the tourist flows between the inner city scenic spots are less sensitive to distance. (4) The primary scenic spots in Nanjing, encompassing the three systems (Sun Yat-sen's Mausoleum system, Confucius Temple system and central business district system), can be abstracted into three models: diffusion from a single center, clustering to a single center and balancing between multiple centers. The flow pattern is characterized by the primary flows between core nodes, secondary flows between the core and minor nodes and between the minor nodes, and the tertiary flows between the minor and normal moves and between the normal nodes.