大陆居民赴台湾自由行旅游流网络分析及演化研究

随着人们旅游消费能力逐渐增强，对旅游品质的要求越来越高，自由行已经成为人们主要出游方式之一。文章以大陆居民赴台湾自由行为对象，采用数据挖掘技术采集我国最大旅游分享社区网站2009年以来的台湾自由行行程信息，构建了细粒度的旅游流有向网络，并使用社会网络理论进行分析。研究表明，该网络属于无标度网络，网络规模为2656，但平均路径只有4.3，网络密度较低，节点的度分布满足幂律分布，网络核心区与边缘区互动较少。按照时间顺序构建旅游流网络，研究网络的演化并分析其演化动力，发现网络规模、连接数量都与行程数量近似于比例增长，网络演化过程中新增加的节点优先链接网络中的明星节点。文章首次从网络演化的角度使用社会网络理论研究旅游流，丰富和拓展了旅游流研究内容。

Network Analysis and Evolutionary Studies Based on Tourist Flows of Mainland Residents's Self-service Traveling in Taiwan

Nowadays, as people's capacity for tourism consumption has been gradually enhanced, the demand for a quality touristic experience is accordingly higher and higher. Maninland residents's self-service travelling has also become one of the main travel modes chosen by tourists.
The network research of tourist flows has caught the attention of many scholars. This paper presents a new method of data acquisition on tourist flows:we used a web-crawler to classify the data of self-service travelling notes in Taiwan on www.mafengwo.cn, which is the biggest tourist community website in China. Data mining technology was used in order to collect the self-service travelling information in Taiwan from the largest tourism sharing community websites since 2009. First of all, our web-crawler analyzed 909 of text pages from the network to define the fields, take down the names of everyday trips in each station (hereinafter referred to as a node), as well as the announced trip time and schedule. The research data presents the following characteristics:data granularity is fine, but the data volume is very large. This paper aims at analyzing this travel data based on social network theory. After a directed network of tourism flows was established, with the help of small world-based and scale-free network models, the evolution of this network was analyzed. In terms of the order of tourists??journey nodes, an asymmetric adjacency matrix was built. The matrix rows and columns represent the different nodes which are marked by tourists. This paper selected five indicators which can represent the main characteristics of a social network, i.e. network size, density, centricity, cohesive subgroups and the core-periphery relation. Centrality analysis was used to analyze the network of tourist flows. The results of this study showed that the network is scale-free, the network size is 2656, but the path of the network is only 4.3. From the brief findings of the model we found that the network density is low, the degree distribution of the nodes fits power-law distribution, and the center-edge interaction of the network is minimal. In the end, we constructed this tourist flows network in chronological order, to research the evolution of the network and to analyze its evolution power.
Within this research, the dynamic of maninland residents's self-service travelling in Taiwan is:the earlier stage of the network growth is based on the push-pull theory, and in the mid- to late stage of growth this growth is based on the push-pull theory along with the network effect. According to the characteristics of the network nodes, especially the distribution features of the extroverted and introverted, the key network node should be taken seriously with the tourism administrative department. They can forecast crises, and guide and diffuse tourist flows in advance. The question of how to control and diffuse tourist flows through the structure of the network properties and dynamic mechanisms will be one of the most important areas for future research. Another noteworthy point is that this article first researched the tourist flows from the perspective of the network evolution using social network theory. This research method not only enriches the research content and research perspective of tourist flows, but also provides much pertinence and significance in its practical applications.