摘要
为解决移动互联网络信息新环境下旅游客运需求预测模型数据来源有限所导致精度不足的问题,以社交网络数据为切入点,通过网络爬虫技术对社交网络中旅游出行相关文本数据进行采集,运用BERT模型的情感分析对社交网络文本型数据进行定量化处理。在传统旅游客运需求预测模型中融入结构化后的社交网络数据,结合天气、节假日状态等变量构建基于梯度提升回归树(GBRT)的旅游客运需求预测模型。最后,以黄山旅游风景区为实例对象,基于景区客运站实际统计数据和社交网络数据,运用上述方法,进行黄山旅游客运需求预测实证分析。结果表明,社交网络数据有助于提升旅游客运需求预测精度,基于社交网络数据的旅游需求预测模型平均预测精度相较于传统模型提升了10.81%。
In order to solve the problem of insufficient accuracy caused by limited data sources of tourism passenger demand forecasting model under the new information environment of mobile internet, the text data related to travel in social networks were collected through web crawler technology, which took social network data as an entry point. Then, the sentiment analysis of BERT model was used for quantitative processing of social network text data. The structured social network data was integrated into the traditional tourism passenger transport demand forecasting model, and the gradient boosting regression tree(GBRT) based tourism passenger transport demand forecasting model was built with the combination of variables such as weather and holiday state. At last, taking Huangshan scenic area as an example, based on the actual statistical data and social network data of the passenger station in the scenic spot, the above methods were used to make an empirical analysis of Huangshan tourist passenger demand forecast. The results show that the social network data is helpful to improve the accuracy of tourism passenger transport demand prediction. The average prediction accuracy of the tourism demand prediction model based on social network data is 10.81% higher than that of the traditional model.
作者
陈坚
彭涛
曹晏诗
刘柯良
CHEN Jian;PENG Tao;CAO Yanshi;LIU Keliang(School of Traffic and Transportation,Chongqing Jiaotong University,Chongqing 400074,China)
出处
《重庆交通大学学报（自然科学版）》
CAS
CSCD
北大核心
2022年第10期41-47,76,共8页
Journal of Chongqing Jiaotong University(Natural Science)
基金
国家社会科学基金西部项目/National Social Science Foundation of China(17XGL009)
重庆市教育委员会科学技术研究计划项目(KJQN202001611)
重庆市教委“成渝地区双城经济圈建设”科技创新项目(KJCXZD2020029)。
关键词
交通运输工程
旅游客运
自然语言处理
GBRT模型
BERT模型
需求预测
traffic and transportation engineering
tourist passenger transport
natural language processing
GBRT model
BERT model
demand forecasting
