abstract
- The following work dissects the popular opinion on the China-Taiwan conflict via a holistic sentiment and emotion analysis on data obtained from the Twitter API, aiming to go beyond a polar interpretation of information with the objective of expanding the current understanding of said subject through the acquisition of a structured numerical representation of the data in order to classify the opinions using a softmax function to evaluate the presence of emotions conveyed in the sample. 5809 tweets related to the recent international tensions were obtained, pursuing keywords like ¿Taiwan¿, ¿China¿ and ¿war¿. First, raw data was analysed by observing the most frequent hashtags and by performing Latent Dirichlet Allocation (LDA) to find possible topics. After that, the dataset was subjected to a sentiment analysis process done with the Twitter-RoBERTa-base model complementing the results with the text2emotion library. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.