Study On The Method Of Forecasting The Number Of Tourists Based On Internet Search Index

With the gradual improvement of per capita income and the rapid promotion of people’s material and cultural needs, traveling is becoming a leisure activity of people when they are free. The amount of tourists was obviously season-related in most areas of China. During the travel season, most of the cities and scenic areas would be in congestion—creating unnecessary friction. It caused a lot of trouble in the management of city and scenic area. When is the rush hour? And how many visitors would be there? Managers are deeply concerned about these issues. Because these would help them formulate effective measures before the occurrence of unnecessary friction. Research suggests that it has a distinct correlation between Internet search data and social economic activities. Baidu is the largest Chinese search engine in the world, and its utilization is very high in terms of domestic Internet search. Baidu Index, which is released by Baidu, can explain the tendency of the internet search volume of some specific keywords. That is to say, Baidu Index is representative. In this paper, according to the six elements of tourism—food, accommodation, transportation, traveling, shopping, entertainment—we selected some keywords related to Beijing tourism, and based on their Baidu Index and the usage of Baidu search engine, we optimized the Internet search data of these keywords, and established the VAR model and BPNN model, which were used to predict the amount of domestic tourists to Beijing. In this paper, we found that there was cointegration relationship between the optimized Internet search index and the amount of domestic tourists to Beijing, while they were not completely linear. In addition, in this paper, we found that the BP neural network model which we established with the optimized Internet search index had good ability in forecasting the amount of domestic tourists to Beijing.