The novel coronavirus-19 pandemic, also known as the COVID-19 pandemic, has continued to affect the global economy in consequential ways as the effects of the virus continue to put companies and industries at risk. The service sector is being faced with unique difficulties caused by severe demand shortages fueled by lost income and uncertainty about the future, particularly for the hospitality and tourism industries due to less traveling (Boone et al., 2020, pp. 39-40). Due to the sudden demand-side shortages, it is especially important for managers to understand the perspectives of those customers in order to best adapt to the changing environment in the hospitality industry.
This empirical study extracts the latent topics from a large training set of tens of thousands of online reviews that were scraped from Agoda.com regarding accommodations in South Korea. In order to analyze qualitative information from text at scale, text mining techniques can be utilized, specifically topic models. Latent Dirichlet allocation (LDA) was first proposed by Blei et al. (2003) as a stochastic, hierarchical Bayesian topic model that learns the latent topics in a corpora of text and assigns probabilities to documents (e.g., reviews) based on the probabilities that those documents contain the topics. LDA has been used by several studies to analyze online reviews of accommodations in order to gain insight into customer experience and decision making processes (Guo et al., 2017; Sutherland et al., 2020; Sutherland and Kiatkawsin, 2020; Zhang, 2019a, 2019b). This study also utilizes latent Dirichlet allocation to evaluate customers eWOM over time.
Using latent Dirichlet allocation, the extracted topics are utilized to evaluate a testing subset of data ranging from pre-COVID-19 to post-COVID-19 reviews. Changes in the importance of topics are detected and analyzed to establish how the COVID-19 pandemic has impact on the topics of discussion in customer online reviews in a Topics Over Time (ToT) approach. The results show that there is a statistically significant increase in the online reviews regarding comments about some accommodation characteristics (e.g., quality of facilities) while there is a significant decrease in other areas of interest.
: 사회과학분야 > 관광학
한국학술정보㈜의 모든 학술 자료는 각 학회 및 기관과 저작권 계약을 통해 제공하고 있습니다.
이에 본 자료를 상업적 이용, 무단 배포 등 불법적으로 이용할 시에는 저작권법 및 관계법령에 따른 책임을 질 수 있습니다.