气候因子的情绪效应

The Emotional Effects of Climatic Factors

研究耦合了逐日气候数据和情绪数据,使用关键词分析技术和多项式逻辑回归模型考察气候与情绪之间的关联。情绪数据创新性地使用微博大数据替代传统的问卷数据,以全国五个典型城市为样本,爬取这五座城市在新浪微博上关于气候内容的文本,通过自然语言处理和机器学习得到情绪数据。研究发现:气候与情绪之间存在着复杂的曲线关系;不同气候因子对情绪的影响力是有差异的,综合考虑关键词TF-IDF权重、显著性指标和偏回归系数权重等,得出气候因子的影响力中日照和湿度对情绪的影响力最大,气温和降水其次,风速最小。

This study integrates daily climate data with emotion data,employing keyword analysis techniques and polynomial logistic regression models to investigate the correlation between climate and emotions.Innovatively,the emotion data is derived from large-scale Weibo data instead of traditional survey data,focusing on five representative cities in China.The study collected textual content related to climate from Sina Weibo posts in these cities.Through natural language processing and machine learning,emotion data was extracted.The research reveals a complex curvilinear relationship between climate and emotions.The impact of different climate factors on emotion varies;taking into account keyword TF-IDF weights,significance indicators,and partial regression coefficient weights,it was found that sunshine and humidity have the greatest impact on emotions,followed by temperature and precipitation,with wind speed having the least impact.