Korean News Corpus for Changes in Child Abuse before and after COVID-19 Outbreak

During the COVID-19 pandemic, the issue of child abuse surged to the forefront of societal concerns in Korea. The public became increasingly attentive to child welfare issues, and every major Korean newspaper covered instances of child abuse, delving into the details of each case and the backgrounds of both victims and perpetrators. To trace the origin of this heightened awareness, we analyzed Korean news articles from the period before and after the pandemic's onset. We sourced articles from July 2017 to June 2022 (spanning two and a half years before and after the outbreak) using six search terms: "child abuse," "child neglect," "corporal punishment," "emotional abuse," "physical abuse," and "sexual abuse" on Naver News, a leading Korean news portal. Our methodology involved Part-Of-Speech (POS) tagging and stop-word removal to assess keyword frequency. Our objective was to determine if the COVID-19 pandemic was a direct catalyst for the uptick in child abuse reports or if the issue had been previously underreported. Additionally, we built dynamic semantic networks from the articles to chart monthly shifts in the narratives around child abuse in Korea and the society's evolving response to it. The gathered news data was archived in Pickle files, categorized by various dictionary keys such as publication date and time, publisher, title, URL, and article content.

在新型冠状病毒肺炎（COVID-19）大流行期间，韩国儿童虐待问题一跃成为社会关注的焦点议题。公众对儿童福利议题的关注度与日俱增，韩国各大主流报纸均对儿童虐待案例进行了报道，深入剖析每起案件的细节以及受害者与施害者的背景情况。为追溯此次关注度提升的根源，本研究对大流行暴发前后的韩国新闻报道展开分析。本研究以韩国头部新闻门户网站Naver新闻为数据源，通过6个检索词——儿童虐待（child abuse）、儿童忽视（child neglect）、体罚（corporal punishment）、情感虐待（emotional abuse）、身体虐待（physical abuse）与性虐待（sexual abuse），采集了2017年7月至2022年6月（即大流行暴发前后各两年半）的新闻报道。本研究采用的方法包括词性标注（Part-Of-Speech (POS) tagging）与停用词移除，以评估关键词的出现频率。本研究的目标在于厘清：新冠大流行是否为儿童虐待报道数量激增的直接诱因，抑或是该议题此前存在报道不足的情况。此外，本研究基于采集的新闻报道构建动态语义网络，以梳理韩国儿童虐待相关议题的叙事脉络月度变化，以及社会对此议题的响应演进历程。本研究采集的新闻数据已归档为Pickle文件，通过发布日期与时间、发布机构、标题、统一资源定位符（URL）以及文章内容等多种字典键进行分类存储。