基于R/S-TDC-EMD-KPCA方法的网络搜索信息提取及游客流量预测——以海南省为例

Web Search Information Extraction and Tourist Flow Forecasting Based on R/S-TDC-EMD-KPCA Method——A Case Study of Hainan

提出了一种网络搜索信息提取及合成新技术——R/S-TDC-EMD-KPCA方法,首先利用重标极差法(R/S)和时差相关法(TDC)选择具有预测能力的关键词,并对关键词搜索量分别进行经验模态分解(EMD)降噪,然后利用核主成分(KPCA)方法合成网络综合搜索指数.以海南省月度游客流量为预测对象进行实证研究,结果表明,融合了网络综合搜索指数的模型在预测精度方面均优于其他基准模型.说明该文提出的方法能高质量的提取和合成网络搜索信息,进而可有效地应用于游客流量的辅助预测.

In this paper,a new web search information extraction and synthesis technology,the R/S-TDC-EMD-KPCA method,is proposed.First,the rescaled range method(R/S)and the time difference correlation(TDC)method are used to select the keywords with predictive ability.The empirical mode decomposition(EMD)noise reduction is performed on the selected keyword search volume,and then the kernel principal component(KPCA)method is used to synthesize the comprehensive network search index.An empirical study is conducted with the monthly tourist flow of Hainan Province as the forecast object.The results show that the model integrated with the network comprehensive search index is better than other benchmark models in terms of prediction accuracy.Therefore,the proposed method can extract and synthesize network search information with high quality,which can be effectively used for auxiliary prediction of tourist flow.