Kim Hyojung, Park Minjung
Department of Fashion Industry, Ewha Womans University, 52, Ewhayeodae-gil, Seodaemun-gu, Seoul 03760, South Korea.
Heliyon. 2023 Jul 13;9(7):e18048. doi: 10.1016/j.heliyon.2023.e18048. eCollection 2023 Jul.
The growth of digital media usage has accelerated the development of big data technology. According to the agenda-setting theory, news media inform the public regarding major agendas and business cycles. This study investigated 168,786 news documents from 2016 to 2020 related the South Korea fashion business using Python. A total of 19 topics were extracted through latent Dirichlet allocation and then transformed into structured data using a time series approach to analyze significant changes in trends. The results indicate that major fashion industry topics include business management strategies to increase sales, diversification of the retail structure, influence of CEOs, and merchandise marketing activities. Thereafter, statistically significant hot and cold topics were derived to identify the shifts in topic themes. This study expands the fashion business contexts with agenda-setting theory through big data time series analyses and can be referenced for the government agencies to support fashion industry policies.
数字媒体使用的增长加速了大数据技术的发展。根据议程设置理论，新闻媒体向公众通报重大议程和商业周期。本研究使用Python调查了2016年至2020年期间与韩国时尚产业相关的168,786篇新闻文档。通过潜在狄利克雷分配提取了总共19个主题，然后使用时间序列方法将其转换为结构化数据，以分析趋势的显著变化。结果表明，主要时尚产业主题包括增加销售额的商业管理策略、零售结构的多元化、首席执行官的影响力以及商品营销活动。此后，得出了具有统计学意义的热门和冷门主题，以识别主题主题的变化。本研究通过大数据时间序列分析，用议程设置理论扩展了时尚商业背景，可为政府机构支持时尚产业政策提供参考。
