COVID-19 Misinformation Detection on Weibo by Support Vector Classifier

The ongoing COVID-19 pandemic has led to a simultaneous infodemic, marked by a surge of accurate and inaccurate information circulating online and across social media platforms. This study is centered on creating a Support Vector Classifier model designed to categorize misinformation about COVID-19 shared on Weibo, a prominent Chinese social media platform. The dataset utilized, known as the CHECKED dataset, is a crucial tool for those engaged in misinformation detection, especially within the Chinese linguistic landscape and in the context of the current global situation. The findings demonstrate that the Support vector machine model crafted for this study demonstrates robust performance across several pivotal metrics, including accuracy, F1-score, recall, and precision, with an accuracy of 89.07%, an F1-score of 87.33%, a recall of 89.07%, and a precision of 90.38%.