Prediction Of Domestic Tourist Flow Based On Baidu Search Data

As a national 4A-level scenic spot in our country,Siguniang Mountain Scenic Area represents the average level of national tourist attractions.Its development not only brings huge economic benefits to the local area,but also has a certain impact on the entire tourism industry and other industries in society.However,both consumers and tourism-related departments are affected by the asymmetry of tourism information,causing most scenic spots to be overloaded with tourists during summer or National Day holidays,which has a certain negative impact on the resources and environment of the scenic spots and hinders the development of various scenic spots ECO development.Therefore,the establishment of an effective passenger flow forecasting model and timely and accurate forecasting of passenger flow in scenic spots will not only help consumers adjust their decision-making behaviors in time before traveling,but also help scenic departments to formulate relevant policies,reduce resource waste,and promote my country’s tourism economy.Healthy and steady development also has important guiding significance.Based on this,this article takes the Siguniang Mountain Scenic Area as an example,and uses the keyword search volume from January 4,2016 to January 20,2020 to predict and analyze the passenger flow of the scenic spot obtained through crawler technology.First select keywords,expand the keyword library through three ways,and use word segmentation software to obtain 9 core keywords from the keyword library.After introducing the search volume of core keywords,through Granger test,it is concluded that Baidu search data can be used to predict the weekly passenger flow of Siguniang Mountain Scenic Area.Then,the time series AR(2)model,support vector regression SVR model and gradient boosting regression GBRT model were constructed to fit and predict the weekly passenger flow of the scenic spot.The GBRT model has the best prediction effect,which is that2 R is 96.76%,and the RMSE value is 0.18.Finally,the timeliness of Baidu search data is studied and the Baidu search data under the lead time is used to fit and predict the weekly passenger flow of the Siguniang Mountain Scenic Area during the summer period in 2019,and combined with my country’s major health epidemic events to explore its four The impact of the Siguniang Mountain Scenic Area.The results show that:(1)Baidu search data can be used to predict the passenger flow of Siguniang mountain scenic spot;(2)The gradient lifting regression model is the best among the three models;(3)Baidu search data one week in advance can effectively improve the prediction accuracy of the model;(4)The prediction results of COVID-19 were high.The above conclusions show that choosing an appropriate model to predict the passenger flow of a scenic spot can not only effectively supplement the statistical data of tourist flow obtained by traditional methods,but also facilitate the timely decision-making of tourists and the relevant departments of the scenic spot or formulate relevant policies to improve the management level of the scenic spot.