突发涉水事件的舆情分析研究  

Study on public opinion analysis of sudden water events

作　　者：成建国[1] 张鸿星[2] 唐彦[3] 王汪送 张进[3] 李晓东[3] 冯钧[3] CHENG Jianguo;ZHANG Hongxing;TANG Yan;WANG Wangsong;ZHANG Jin;LI Xiaodong;FENG Jun(Water Resources Information Center, the Ministry of Water Resources, Beijing 100053, China;Water Resources Division, the Ministry of Water Resources, Beijing 100053, China;College of Computer and Information, Hohai University, Nanjing 211100, China)

机构地区：[1]水利部水利信息中心,北京100053 [2]水利部水资源司,北京100053 [3]河海大学计算机与信息学院,江苏南京210098

出　　处：《水利信息化》2017年第4期21-27,共7页Water Resources Informatization

基　　金：国家自然科学基金项目(61370091)

摘　　要：互联网信息公开、迅速、透明且易于收集,对于突发性水利事件的分析,互联网舆情大数据分析是一个发展趋势,这种方式能以较小代价快速获取数据,分析得到有价值的信息,正成为辅助常规水利事件监测处理的有力方式。在研究和应用现状分析的基础上,通过对搜索引擎、社交论坛和新闻报道等渠道采集相关数据并分析挖掘,提出面向突发事件的热度和情感分析模型,并对2017年嘉陵江铊污染突发事件,从9个情感维度进行分析。微博用户的陈述、愤怒、担心的比重较大,论坛评论的用户大部分都表示出较理性的心态,新闻评论的网友多处于担心和追责状态,负面情绪偏多。突发涉水事件的舆情传播很快,因为事件应急处理措施快速生效,水污染情况得到有效控制,事件舆情逐渐趋于稳定。Internet information disclosure is transparency and easy to collect.For the analysis of sudden water events,Internet public opinion analysis is a large trend of development.This analysis can be very small input to obtain valuable information,and get valuable result,is becoming an indispensable way to assist in the routine monitoring of water events.On the basis of analyzing the research and application situation,through search engine,micro-blog,forums and Internet news reports big data related to water incident event,the article proposes an event heat analysis model and emotional analysis model.Both models are used for discovering the public opinion of the Jialing River thallium pollution in2017.It is analyzed from nine emotional dimension,is discovered that statement,anger and worrisome are the main emotions in micro-blog,the forum users expresses a rather rational emotion,while the news commenters are more negative,expressing a strong desire to find the people and organization accountable for this incident.The public sentiment become stable since the incident is emergency disposal plan works effectively.

关 键 词：突发性水利事件 大数据挖掘 舆情 事件热度分析 情感分析 

分 类 号：P426.616[天文地球—大气科学及气象学]