Application of Sentiment Analysis in Translanguaging and Translation

The purpose of this article is to emphasize the importance of sentiment analysis in translanguaging and translation. In sentiment analysis, “sentiment” refers not only to emotional feelings of love or hate, but also to opinions or attitudes. For this purpose, this study collected and examined three sets of original English transcripts and their Korean translations using the TED Corpus Search Engine (TCSE), developed by Yoichiro Hasebe at Doshisha University in Japan. With the data sets, this study performed BERT-based multilingual sentiment analysis on Google CoLab. The results revealed that sentence length and conjunction usage did not influence sentiment scores, whereas emotional or attitudinal words and phrases did affect the scores. These findings can lead to the development of new pedagogies and practical applications in the field of translation education. In addition, highlighting the significance of coherence and implicature, this study seeks to find the pragmatic or sentimental equivalence in the digital learning landscape.