基于情感挖掘和话题分析的旅游舆情危机演化特征——以“丽江女游客被打”事件为例

The Evolution Characteristics of Tourism Opinion Crisis based on Emotion Mining and Topic Analysis:A Case Study of the Incident of "Female Tourist Attacked in Lijiang"

现有旅游业危机管理体系缺乏对旅游舆情危机的关注。该研究聚焦旅游业的舆情危机,从情感传播的视角对旅游舆情危机的演化规律和行为特征进行深入剖析,尝试探寻旅游舆情危机传播演化的一般性规律。文章以"丽江女游客被打"事件为例,采用网页数据采集工具爬取互联网中新闻报道和用户评论的相关数据,通过提取危机关键词、识别舆情话题、计算情感强度,并结合危机生命周期的阶段性特征,对旅游舆情危机进行综合分析。研究结果显示:(1)公众心理存在"放大效应",对危机事件的评价更容易被舆论场中最突出的行为特征所掩盖,而忽视事件自身客观规律,导致以点概面的评价结果;(2)事件引发的舆情话题存在纵深发展和横向扩散两种演化方式,并衍生出新的话题,新的舆情话题伴随着负面情绪的波峰出现;(3)由公共安全引发的旅游舆情危机突发且易逝,事件生命周期长度具有显著的不确定性,与现实中案情调查进展、政府应对及信息公布和突发新闻等影响因素密切相关。

The popularity of mobile social network platforms, such as Sina Weibo, We Chat, etc., have substantially changed our way of living and communication. As an area closely connected with Internet news media, tourism industry is facing with unprecedented opportunities as well as challenges. Tourism opinion crisis events under the environment of Internet, such as "Fenghuang(Phoenix) Ancient Town charges for residents", "Qingdao shrimps with high price", and "Female tourist attacked in Lijiang",etc., have raised widespread attention with the facilitation of online social network platforms, and have made a great impact on the image of tourist spots and the development of tourist activities. However,tourism opinion crisis, as the outcome of the development process of platform economy virtualization,has received less attention in the study of tourism crisis management. To amend this gap of knowledge,this implements the following research.First, a novel classification rule for tourism crisis is proposed, on the basis of combing domestic and foreign literature concerning tourism crisis management and network public opinion. The concept of tourism opinion crisis, which is the research subject in this paper, is expounded from the perspective of network public opinion. Taking the"Female tourist attacked in Lijiang"incident as an example, the webpage data crawling tool is used to extract relevant news reports and user comments data from Sina Weibo. And the final event-related corpus is obtained after data preprocessing.Second, through crisis keywords extraction, lyric topics identification, and emotional intensity calculation, a comprehensive analysis was carried out with respect to the life cycle of the tourism opinion crisis. It is found that:(1) There exists an "amplification effect" in public psychology. Public evaluation of crisis events is more easily leaded by the most prominent behavioral characteristics of public opinion and it usually ignores the universal evolution law of the events, which results in onesided evaluation results.(2) Two evolution patterns of public opinion topics are found, i.e. deeply developing and transverse diffusing. New deriving topics are often accompanied by peaks of negative emotions.(3) The tourism opinion crisis caused by public security is emergent and perishable, and the life cycle length of crisis events has significant uncertainty.Lastly, some targeted suggestions are proposed for tourism crisis management. The research not only reveals the dissemination characteristics of tourism opinion crisis under the environment of Internet,but also offers valuable insights for the supervision departments to improve public opinion management.