Research On Forecast Of Tourist Volume Based On Web Search Key Words

With the booming Internet age,the Internet has become an inseparable part of our daily lives.When making consumer decisions,people are accustomed to searching for relevant information online,based on these information resources,to assist themselves in making consumption decisions.At present,the living standards of our people have gradually increased,the disposable income has increased substantially,and the demand for tourism has become more and more important.Tourism has become an increasingly popular choice for people's leisure time.On holidays,the number of tourists will skyrocket,and the traditional tourism decision-making model can no longer meet people's needs.Consumers use search engines as a platform,search keywords as the key,and find travel information on the Internet.This has become a fashion.Therefore,this paper takes Sanya as an example to analyze the relationship between consumer network search keyword data and the amount of tourists,and establish a corresponding statistical model to predict the amount of tourists.Based on the tourism resources in Sanya,this paper first selects 15 keywords by using technical selection,subjective selection,range selection and keyword expansion.It uses gray correlation analysis and Spearman correlation coefficient.From the quantitative point of view,the correlation between the web search data of these 15 keywords and the Sanya visitor volume data is analyzed.Based on the correlation results,the principal component analysis is performed on 15 keyword data,and 5 principal components are extracted,and then according to Based on the five principal component data and the Sanya visitor volume data,BP neural network prediction model,support vector machine prediction model and random forest prediction model were constructed.Based on the prediction results of these three models,the data of the tourists was established.Multiple regression comprehensive prediction model.Finally,the comparison chart between the predicted data and the actual data is given.From the qualitative and quantitative aspects,the prediction ability and actual application effect of each model are evaluated.The conclusions drawn in this paper are as follows: First,the search data of the finalized network search keywords are related to the Sanya visitor volume data.Second,among thethree prediction models constructed,BP neural network has the best prediction effect,but the support vector machine model prediction is more practical than BP neural network.Third,the prediction effect of the multiple regression comprehensive prediction model is better than the other three single prediction models.