Pathak Tanmay S, Athavale Harsh, Pathak Amey S, Athavale Sunita
Electronics and Communication Engineering, International Institute of Information Technology Hyderabad (IIITH), Hyderabad, IND.
Computer Science and Engineering, Manipal University Jaipur, Jaipur, IND.
Cureus. 2021 Oct 30;13(10):e19141. doi: 10.7759/cureus.19141. eCollection 2021 Oct.
Introduction The World Health Organization (WHO) is a specialized agency of the United Nations responsible for international public health. Established on April 7, 1948, it has since played a pivotal role in several public health achievements and has had considerable success. But never since the establishment of the WHO has it faced a pandemic of such a huge scale. The spread of the coronavirus and the inability of the WHO to contain it has raised many questions about its efficiency and role. The present study explores the range of emotions and sentiments evoked by public health information posts of WHO over the course of the pandemic. Methods This study uses Bidirectional Encoder Representations from Transformers (BERT), which is a neural network-based technique for natural language processing. Three timeframes of five months each, starting from March 2020, were defined. A total of six posts, two posts from each timeframe, were then analysed. Comments were classified as positive, neutral and negative. The broader positive and negative classes were further subclassified into two classes each. Natural language processing was further applied to obtain results. Results The general trend of the sentiments over the period of pandemic showed a significant and dominant proportion of negative comments that overshadowed the neutral, positive and irrelevant comments over all timeframes. Specifically, the negative sentiments peaked during the second timeframe. The negativity was directed more towards the WHO, governments and people not complying with coronavirus disease 2019-appropriate norms. Positive comments were mostly expressed towards health workers. Conclusion An unusually high proportion of negative sentiment was observed in response to relatively innocuous public health posts. This may be a result of heightened anxiety, questionable credibility of the sources of information and geopolitical power play maligning the image of the WHO.
引言 世界卫生组织(WHO)是联合国负责国际公共卫生的专门机构。它于1948年4月7日成立,自那时起就在多项公共卫生成就中发挥了关键作用,并取得了相当大的成功。但自世界卫生组织成立以来,从未面临过如此大规模的大流行。冠状病毒的传播以及世界卫生组织无法遏制它,引发了许多关于其效率和作用的问题。本研究探讨了在大流行期间,世界卫生组织的公共卫生信息帖子所引发的一系列情绪和情感。
方法 本研究使用了基于神经网络的自然语言处理技术——来自变换器的双向编码器表征(BERT)。定义了从2020年3月开始的三个各为期五个月的时间框架。然后分析了总共六篇帖子,每个时间框架两篇。评论被分类为积极、中性和消极。更宽泛的积极和消极类别又各自进一步细分为两个类别。进一步应用自然语言处理以获得结果。
结果 在大流行期间,情绪的总体趋势显示,负面评论占显著且主导的比例,在所有时间框架内都超过了中性、积极和无关评论。具体而言,负面情绪在第二个时间框架达到峰值。负面情绪更多地指向世界卫生组织、政府以及不遵守2019冠状病毒病相关规范的民众。积极评论大多是针对卫生工作者表达的。
结论 对于相对无害的公共卫生帖子,观察到异常高比例的负面情绪。这可能是焦虑加剧、信息来源可信度存疑以及地缘政治权力博弈损害世界卫生组织形象的结果。