摘要
本研究基于百度指数数据,构建旅游需求预测模型,旨在优化旅游资源配置并提升预测精度。通过灰色关联度分析,发现6个核心关键词与大连市游客人数的关联度系数均超过0.75,表明网络搜索数据与旅游需求存在显著正向关联。进一步通过Spearman相关性分析,从200个候选词中筛选出12个强关联关键词,其中“景点”类关键词占比较高,凸显了大连丰富的景观资源对旅游需求的核心推动作用。模型对比结果显示,传统岭回归和Lasso回归的预测精度显著高于PCA改进模型,表明主成分分析虽能缓解共线性问题,但可能因信息损失而削弱模型对高维数据的理解能力。研究还发现,节假日和寒暑假期间游客的决策模式存在差异,法定节假日的即时决策导致实际游客量超出预测值,而暑假期间的前置规划则使预测值高于实际游客量。研究结论为旅游管理部门提供了优化资源配置和精准营销的实证依据,也为基于网络搜索数据的旅游需求预测提供了新的视角和方法。
Based on Baidu index data,this study constructed a prediction model of tourism demand in Dalian,aiming to optimize the allocation of tourism resources and improve the prediction accuracy.Through gray correlation analysis,it was found that the correlation coefficients between six core keywords and the number of tourists in Dalian were over 0.75,indicating that there is a significant positive correlation between online search data and tourism demand.Through Spearman's correlation analysis,12 strong keywords are selected from 200 candidate words,among which the proportion of“attractions”keywords reaches 66.7%,which highlights the core driving effect of Dalian's rich landscape resources on tourism demand.Comparison of the models shows that the prediction accuracy of traditional ridge regression and Lasso regression is significantly higher than that of the improved version of PCA,which indicates that although principal component analysis can alleviate the problem of covariance,the loss of information may weaken the model's ability to understand high-dimensional data.The study also found differences in the decision-making patterns of tourists during holidays and summer and winter vacations:instantaneous decision-making during legal holidays resulted in actual tourist arrivals exceeding the predicted values,whereas front-loaded planning during summer vacations resulted in predicted values higher than actual tourist arrivals.These findings provide a scientific basis for tourism management to optimize resource allocation and precise marketing,as well as a new perspective and methodology for tourism demand forecasting based on web search data.
作者
李晓菲
郭小婉
Xiaofei Li;Xiaowan Guo(School of Statistics,Dongbei University of Finance and Economics,Dalian,Liaoning,China;School of Information and Business Administration,Dalian Neusoft University of Information,Dalian,Liaoning,China)
基金
辽宁省社会科学规划基金项目青年基金项目:辽宁省旅游形象的短视频传播力评价与提升策略研究(L24CTJ001)。
关键词
百度指数
旅游需求预测
机器学习算法
Baidu Index
Tourism Demand Prediction
Machine Learning Algorithms
