Self-disclosure on Twitter During the COVID-19 Pandemic: A Network Perspective
Self-disclosure on Twitter During the COVID-19 Pandemic: A Network Perspective
复制标题
DOI：
10.1007/978-3-030-86514-6_17
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Prasanna Umar;Chandan Akiti;A. Squicciarini;S. Rajtmajer
中科院分区：
文献类型：
--
作者：
Prasanna Umar;Chandan Akiti;A. Squicciarini;S. Rajtmajer
在COVID-19大流行造成的社交距离、隔离和日常干扰中，用户在在线社交媒体上的活跃度增加，为自我披露提供了更多机会。我们研究了自我表露的发生率和演变时间，因为重要的事件在整个流行病的时间轴展开。使用基于BERT的监督学习方法，我们将超过3100万条COVID-19相关推文的数据集标记为自我披露。我们映射用户的自我披露模式，描述个人的启示，并检查用户的披露在不断发展的回复网络。我们采用自然语言处理模型和社交网络分析来研究用户在2019冠状病毒病大流行期间寻求社交联系和聚焦对话时的互动网络中的自我披露模式。我们的分析显示，在世界卫生组织于2020年3月11日宣布全球大流行之后，推文中的自我披露水平有所提高。我们解开网络层次的自我表露模式，并显示自我表露的特点是时间持久的社会联系。我们认为，在追求社会回报的用户故意自我披露，并与类似披露用户。最后，我们的工作表明，在这种追求中，用户可能会披露私密的个人健康信息，如个人疾病和潜在的隐私风险。
Amidst social distancing, quarantines, and everyday disruptions caused by the COVID-19 pandemic, users’ heightened activity on online social media has provided enhanced opportunities for self-disclosure. We study the incidence and the evolution of self-disclosure temporally as important events unfold throughout the pandemic’s timeline. Using a BERT-based supervised learning approach, we label a dataset of over 31 million COVID-19 related tweets for self-disclosure. We map users’ self-disclosure patterns, characterize personal revelations, and examine users’ disclosures within evolving reply networks. We employ natural language processing models and social network analyses to investigate self-disclosure patterns in users’ interaction networks as they seek social connectedness and focused conversations during COVID-19 pandemic. Our analyses show heightened self-disclosure levels in tweets following the World Health Organization’s declaration of pandemic worldwide on March 11, 2020. We disentangle network-level patterns of self-disclosure and show how self-disclosure characterizes temporally persistent social connections. We argue that in pursuit of social rewards users intentionally self-disclose and associate with similarly disclosing users. Finally, our work illustrates that in this pursuit users may disclose intimate personal health information such as personal ailments and underlying conditions which pose privacy risks.