Understanding Tourist Perspectives of Borobudur through LDA Topic Modelling Using Elbow Method

The primary objective of this research is to unveil the underlying themes and topics within tourist reviews that evaluate the Borobudur temple complex and its surroundings. This goal will be achieved through the application of Latent Dirichlet Allocation (LDA) topic modeling. The method involves collecting English-language reviews from the online platform TripAdvisor and subsequently subjecting them to preprocessing techniques to ensure proper data formatting for subsequent analysis. The optimal number of topics is determined using the elbow method within the Optimize Parameters (Grid) Operator in the RapidMiner platform. The preprocessed data is then subjected to analysis utilizing the Latent Dirichlet Allocation (LDA) model, which enables the extraction and identification of distinct topics. The results reveal a diverse range of topics pertinent to tourists’ experiences, encompassing accessibility, prominent features, guided tours, and activities during sunrise and sunset. These findings offer valuable insights into the key aspects emphasized by tourists and their corresponding sentiments. Such insights can be harnessed to enhance tourism management and services within the Borobudur temple complex.