Detecting fake news on social networks via linguistic features and information-seeking patterns during the Covid-19 period

News spreads most widely on social media; however, these platforms can also disseminate fake news, especially in times of crisis, leading to serious societal consequences. To understand why people share fake news on social media, we examine information-seeking behavior during the COVID-19 pandemic. Social media platforms are using third parties to verify news articles due to the abundance of fake news, which is increasingly time- and resource-intensive. In response, we propose a news representation model that leverages information-seeking patterns to extract syntactic, emotional, lexical diversity, and readability features from social media posts for accurate fake news detection. We demonstrate that the proposed model significantly improves detection performance over traditional language representation models, validated using three social media datasets from the COVID-19 pandemic. Our findings indicate that incorporating information search behavior and linguistic features effectively enhances the detection of fake news on social media.