基于向量自回归模型的旅游热门景点预测方法研究
2020-08-04任婕
任婕



摘 要: 常规的预测方法变量之间的皮尔逊相关值较低,造成预测的结果出现偏差,为此,设计基于向量自回归模型的旅游热门景点预测方法。综合不同的搜索引擎数据,计算旅游景点关键词网络搜索指数,对其进行预处理并筛选出与旅游景点热度相关性较强的关键词,利用向量自回归模型对变量进行均值化处理,确定影响最大的网络搜索指数,实现对旅游热门景点的预测。实验结果表明:与常规的灰度预测方法和SVR模型预测方法相比,基于向量自回归模型的预测方法的皮尔逊相关值能够保持在0.8~1.0之间,变量之间具有极强的相干性,适合应用在旅游热门景点预测中。
关键词: 旅游热门景点预测; VAR模型; 关键词搜索指数; 皮尔逊相关系数; 搜索指数计算; 客流量预测
中图分类号: TN911.1?34; TP181 文献标识码: A 文章編号: 1004?373X(2020)03?0158?04
Research on popular tourist spot prediction method
based on vector auto regression model
REN Jie
(Ningxia Academy of Social Sciences, Yinchuan 750021, China)
Abstract: Since the Pearson′s correlation coefficient between the variables in conventional prediction methods is low, which leads to the deviation of predicted results, a popular tourist spot prediction method based on vector auto regression (VAR) model is designed. In combination with the data of different search engines, the internet search indexes of tourist spot keywords are calculated. The keywords are preprocessed, and those with strong correlation for tourist hot spots are screened out. The vector auto regression model is used to average the variables and determine the most influential internet search index to predict the popular tourist spots. The experimental results show that, in comparison with the conventional gray prediction method and the SVR model based prediction method, the Pearson′s correlation coefficient of the prediction method based on vector auto regression model keeps in the range of 0.8~1.0, and the variables are of strong coherence. Therefore, the proposed method is suitable for the prediction of popular tourist spots.
Keywords: popular tourist spot prediction; VAR model; keyword search index; Pearson′s correlation coefficient; search index calculation; tourist flow prediction
0 引 言
随着经济的高速发展,现代人们的生活水平不断提高,人们开始追求更高层次的体验,比如旅游。旅游行业是一个综合性的新兴行业,其突出的特点就是投入成本较少,经济价值回报较高,能够满足人们对精神层面上的需求[1]。一般旅游业具有综合性和季节性特性,可以带动文化教育业、交通运输业以及服务行业等的发展,因此应重视旅游业的发展[2]。
旅游景区因旅游人数的迅速递增,带来了可观经济效益的同时,也考验旅游景区的科学管理能力。根据大量旅游产品的调查资料,对旅游游客流量的走势进行分析,利用合理的方法对旅游热门景点进行预测[3]。……
