【摘要】 以人民日报发布的“寒潮预警”相关新浪微博话题为切入点，利用网络爬虫技术抓取微博用户评论数据；利用LDA和ArcGIS分析新媒体用户在寒潮灾害预警下的情感响应特征与时空分异，为公众对气象灾害的响应研究做出贡献。结果表明：(1)寒潮灾害预警下，新媒体用户表现出“贡献力量”、“不惧寒潮”、“积极乐观”、“关心家乡”等正面情感响应特征以及“抱怨预警”、“瑟瑟发抖”、“出门困难”等负面情感响应特征；(2)各地对寒潮预警话题的响应整体上呈东高西低的趋势，一方面受人口空间分布影响，另一方面降水量丰富的地区响应热度较高；(3)正面评论聚集于东南部地区，负面评论聚集于东北部地区，均在西南部地区发散；(4)随着寒潮预警级别的升高，新媒体用户响应热度升高，情感响应更消极，正面评论集聚性降低，负面评论逐渐向东中部地区发散。更多还原

【Abstract】 Based on the topic of Sina microblog about cold wave warning published by People’s Daily, we used Web crawler technology to capture the comment data of microblog users, and used LDA and ArcGIS to analyze the emotional response characteristics and spatio-temporal differentiation of new media users to cold wave disaster warning. The results show that(1)under the cold wave disaster warning, the new media users show positive emotional response characteristics of“contribution”,“not afraid of cold wave”,“optimistic and cheerful”,“care about hometown”and negative emotional response characteristics of“complaint warning”,“shivering”and“difficult to go out”.(2)On the whole, the response of all localities to the topic of cold wave warning shows a trend of high in the east and low in the west. On the one hand, it is affected by the spatial distribution of population, on the other hand, the areas with abundant precipitation have higher response heat.(3)Positive comments gather in the southeast and diverge in the southwest, while negative comments gather in the northeast and diverge in the southwest, with significant negative hot spots in the northeast.(4)With the increase of cold wave warning level, the response of new media users increases, and the emotional response is more negative. The agglomeration of positive response decreased, and the negative comments gradually spread to the east and central region.更多还原