<![CDATA[X is a digital platform that facilitates the sharing of thoughts and criticisms through written content. A multitude of individuals and organizations depend on the perspectives or sentiments of the general populace while making decisions. Consumers generally rely on the viewpoints of other consumers when it comes to evaluating a product or service that they come across on social media sites. Through the surveillance of social media activity, companies that sell products and services can gain insight into the emotion expressed by consumers towards their offerings. Nevertheless, due to the limitations of writing, which lacks the ability to transmit nonverbal cues like gestures, facial expressions, and intonation, it is often challenging to identify implicit signs such as sarcasm. Sarcasm in a tweet can lead to an erroneous evaluation of the message's sentiment. Hence, it is crucial to conduct sarcasm detection, as it can greatly enhance the outcomes of sentiment analysis. This study assesses the efficacy of four transformer models, namely IndoBERT, RoBERTa, BERT, and BERT Multilingual, in detecting sarcasm in Indonesian on X platform. The experimental results demonstrate that the IndoBERT model, which has been specifically tailored for the task, gets an impressive F1-score of 95%.]]>
Copyrights © 2024
