Conference Paper Published

Abstract

This study investigates the diachronic changes in the sentiments and metaphorical frames in a corpus of news discourse on COVID-19 as a case study to examine the potential implication for applying sentiment analysis in corpus data and its interaction with metaphorical framing changes over time. The corpus contains COVID-19 news articles covering the entire cycle of the pandemic from 2019 to 2024 in Hong Kong. The sentiment analysis of the corpus was explicitly presented. We found that the sentiments of the news are overall objective and slightly positive. The diachronic changes and the interaction between the sentiments and metaphor polarities were discussed with empirical examples from the corpus, aiming to establish an operational approach for exploring the connection between metaphor polarities and the sentiments in large-scale of discourse data.