Analysis Of Jiuzhaigou Network Attention And Construction Of Forecast Model

In recent years, with the rapid development of China’s tourism industry, the major tourist cities and tourist attractions often become overcrowding. This phenomenon not only bring serious impact to the experience of visitors, but also bring the threat to the personal safety of tourists. The current area usually take measures in or after the event, these measures are generally with emergency color, uncertainty, and will bring many follow-up questions. Especially, the measures will bring great inconvenience to tourists. Tourists often had to temporarily change their travel itinerary, the subsequent emotional calm will very difficult. So the research objective of this paper is to take Jiuzhaigou as an example, to build a passenger flow forecasting model of Jiuzhaigou, so as to provide the "ex ante" data support for the scenic area managers and tourists. The main conclusions of this paper are as follows:First, based on the Internet era characteristics of tourists, this article will analyze the tourists’ information search behavior and construction of tourism decision-making model in the Internet Age. Tourist search information throughout the whole process of the whole tourism activities. In the age of the Internet, it is common for people via the Internet search travel information.Second, this paper taking Jiuzhaigou as an example, through the analysis of Jiuzhaigou actual traffic and network attention degree, found the tourist flow of Jiuzhaigou and network attention degree showed almost the same change trend during the year, showing a double peak shape. Around the rest days, holidays and working time, they show regular changes. And found that the Jiuzhaigou network attention will peak before the peak of the passenger flow, the Jiuzhaigou network attention showed a significant effect of the precursor effect to the actual passenger flow.Third, this paper use Jiuzhaigou actual traffic and related Baidu Index as the basic data by ADLM, and build a passenger flow forecasting model of Jiuzhaigou by 2 years, 1year and six months of sample period. And the predictive ability of the model was verified, with respect to the Jiuzhaigou’s official website announced the forecast data, the model significantly increased the prediction accuracy. It is proved that the passenger flow forecasting model of Jiuzhaigou has good predictive ability. And compared the different sample period to construct the model predictions and found 1 year sample period to construct the prediction model have the more predictive ability, that use Baidu Index to build passenger flow forecasting model is not collected data covering the time the longer the better. This is because China is still in the stage of rapid development of the Internet, the proportion of tourists through online access to travel information and search methods are constantly changing. If the sample period span is too long, it will lead to the model of the fitting effect is very good, but the forecast effect is likely to be greatly reduced.