engleski

Topic modelling of Croatian five-star hotel brands posts on Facebook using Latent Dirichlet Allocation

Official websites of hotel brands on Facebook are still an efficient way to interact with their customers, as the interaction between them covers different types of valuable information and benefits. The purpose of this study is to investigate the activities of hotel brands in social media and extract latent posting topics from Croatian five- star hotel brands. The authors used descriptive statistical analysis to analyse hotel brands’ activity and the probabilistic generative topic model Latent Dirichlet Allocation (LDA) to extract latent topics from the large number of hotel brands’ posts on Facebook. The results show that Croatian five-star hotel brands pursue the most common social media marketing objectives based on relationship building, publicity or message strategies, brand building, market research and sales promotion. The paper discusses the application of the Latent Dirichlet Allocation analysis and offers recommendations, challenges, and current insights into applied Croatian marketing strategies on the social media platform.

tourism research, hospitality industry, social media, topic modelling, Latent Dirichlet Allocation (LDA)

nije evidentirano