基于多语言搜索引擎数据的旅游需求预测研究  

Forecasting Tourism Demand with Multi-Language Search Engine Data

作　　者：于翠翠 寇红红 孙少龙 YU Cuicui;KOU Honghong;SUN Shaolong(School of Economics and Management,Xidian University,Xi'an 710126;School of KoU Honghongl Management,Xian Jiaotong University,Xi'an 710049)

机构地区：[1]西安电子科技大学经济与管理学院,西安710126 [2]西安交通大学管理学院,西安710049

出　　处：《系统科学与数学》2022年第9期2383-2398,共16页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金(72101197);中央高校基本科研业务费(SK2021007);陕西省软科学研究计划(2022KRM015)资助课题

摘　　要：在大数据时代,搜索引擎数据(search engine data,i.e.,SED)作为一种强大的工具被广泛用于预测领域,如旅游需求和原油价格预测.值得探究的是,基于不同搜索语言的SED可能有不同的预测能力,引入多语言SED以提高预测目标的预测精度则成为预测领域的研究方向之一.因此,文章采用不同语言搜索引擎数据和人工智能技术,探讨结合多语言SED和单一语言SED在旅游市场需求方面的预测能力.具体地,以中国香港入境游客量为研究样本,以中文,英文,日文和韩文为关键词筛选语言,相应地收集多语言SED,实现1-3期提前预测.实证结果得到3个显著性结果:首先,单一语言SED可以明显提高预测精度.其次,多语言SED也可以作为具有预测能力的数据源,以提高预测精度.第三,多语言SED在各方面的预测能力与单一语言SED预测能力相近,并未表现出明显优势.研究结果可为预测变量的拓展提供理论支撑,以及为旅游业研究人员和从业人员提高旅游需求预测准确度提供参考.In big data era,the tourism related search engine data(SED)as a powerful tool has been widely used in the field of forecast,such as tourism demand for travel destinations and oil price.However,the SED based on different search languages may have different forecasting capability to improve the accuracy.Therefore,this study employed the different language SED with artificial intelligence techniques for exploring the corresponding forecasting power of aggregated multi-language SED and the single different language SED.With Hongkong,China as a research sample,empirical results show three significant results.Firstly,the results reveal that SED in a single language can obviously improve the forecasting accuracy.Secondly,multi-language SED can also be used as a predictor for improving forecasting accuracy.Thirdly,the forecasting power of multi-language SED is no better than that of a single language SED in all horizons.The research results with one-,two-and three-step-ahead forecasting can provide theoretical support for the expansion of prediction variables,and provide reference for tourism researchers and practitcioners to improve the accuracy of tourism demand forecasting.

关 键 词：大数据 搜索引擎数据 机器学习 旅游需求预测 RVFL 

分 类 号：F592.7[经济管理—旅游管理] TP391.1[经济管理—产业经济]