Online sentiment towards iconic species

Online sentiment towards iconic species
复制标题

DOI:
10.1016/j.biocon.2019.108289
复制
发表时间:
2020-01-01
影响因子:
5.9
通讯作者:
Di Minin, Enrico
Di Minin, Enrico
中科院分区:
环境科学与生态学1区
文献类型:
--
作者:
Fink, Christoph;Hausmann, Anna;Di Minin, Enrico

文献摘要

被引文献

相似文献

评估在线公众对生物多样性保护情绪的研究几乎不存在。在保护科学中,社交媒体数据和其他在线数据来源的使用正在增加。我们收集了与犀牛有关的社交媒体和在线新闻数据,犀牛是特别受到非法野生动物贸易威胁的标志性物种,并使用自然语言处理方法评估了对这些物种的在线情绪。我们还使用了离群值检测技术来识别这些数据中最突出的保护相关事件。我们发现,悲剧性事件,如2018年3月苏丹最后一只雄性北方白色犀牛的死亡,引发了最强烈的反应,这些反应似乎集中在西方国家,在犀牛范围之外的国家。我们还发现,社交媒体数据量和在线新闻量之间存在很强的时间交叉相关性,而其他事件只出现在社交媒体或在线新闻中。我们的研究结果强调,公众关注生物多样性的丧失,这反过来又可以用来增加决策者的压力,制定适当的保护行动,可以帮助扭转生物多样性危机。所提出的方法和分析可用于从数字媒体数据中推断对任何生物多样性主题的情感,并检测哪些事件对公众来说最重要。
Studies assessing online public sentiment towards biodiversity conservation are almost non-existent. The use of social media data and other online data sources is increasing in conservation science. We collected social media and online news data pertaining to rhinoceros, which are iconic species especially threatened by illegal wildlife trade, and assessed online sentiment towards these species using natural language processing methods. We also used an outlier detection technique to identify the most prominent conservation-related events imprinted into this data. We found that tragic events, such as the death of the last male northern white rhinoceros, Sudan, in March 2018, triggered the strongest reactions, which appeared to be concentrated in western countries, outside rhinoceros range states. We also found a strong temporal cross-correlation between social media data volume and online news volume in relation to tragic events, while other events only appeared in either social media or online news. Our results highlight that the public is concerned about biodiversity loss and this, in turn, can be used to increase pressure on decision makers to develop adequate conservation actions that can help reverse the biodiversity crisis. The proposed methods and analyses can be used to infer sentiment towards any biodiversity topic from digital media data, and to detect which events are perceived most important to the public.