基于旅游者网络关注度的旅游景区日游客量预测研究——以不同客户端百度指数为例

网络搜索引擎是旅游者获取旅游信息的最重要入口，百度指数通过反映关键词被搜索的次数表征旅游者的网络关注度。文章以三清山为例，首先利用协整理论及格兰杰因果检验分析了PC端和移动端百度指数与实际游客量之间的关系，进一步建立日游客量ARMA模型和分别加入PC 端和移动端百度指数的VAR模型，对游客量预测结果及预测能力进行比较分析，以期通过不同客户端、不同搜索关键词来填补游客量预测过程中旅游网络数据提取的单一性，得到更好的预测效果。发现移动端比PC端百度指数模型具有更好的预测效果，移动端比PC端百度指数对实际游客量的变动具有更好的解释能力。

FORECAST OF TOURISM FLOW VOLUME OF TOURISTATTRACTION BASED ON DEGREE OF TOURISTATTENTION OF TRAVEL NETWORK: A CASE STUDY OF BAIDU INDEX OF DIFFERENT CLIENTS

It is the web search engine that is an important way for tourists to get travel information. Therefore, it is easy to record behavior of tourists. Baidu Index, by searching times of relevant keywords, is able to find out the degree of tourist attention of travel network easily. With the changes of different clients, Baidu Index shows certain spatiotemporal difference and precursor effect. In order to find out the relationships between Baidu Index of PC client and mobile client and the actual visitor number of Mount Sanqingshan, paper used the econometric cointegration theory and Granger causality test. In addition, to forecast Tourism Flow Volume, the paper further establishes ARMA model of the daily visitor number of Mount Sanqingshan and VAR models which add Baidu Index of PC client or mobile client respectively. It is found that:1) There are long-term equilibrium relationships between the actual visitor number of Mount Sanqingshan and Baidu Index of PC client and mobile client of multigroup search keywords; 2) The results of variables Granger causality tests between Baidu Index of PC client and mobile client of Different search keywords and the actual visitor number of Mount Sanqingshan present significant inconsistencies; 3) Among the three prediction model, VAR model of mobile client is of the best prediction accuracy and ARMA model of the daily visitor number of Mount Sanqingshan is of the lowest prediction accuracy.