News Credibility Identification on Facebook

The credibility of information on the World Wide Web (WWW), particularly on Social Networking Sites (SNS), remains a significant concern due to the ease of information sharing without editorial oversight. Online information can often be non-credible, including false, outdated, or biased content. Novice web users frequently struggle to assess the credibility of information shared on SNS, underscoring the need for effective evaluation mechanisms. This paper proposes a news credibility identification framework designed to evaluate web news based on three key credibility categories: correctness, authority, and professionalism. These categories are assessed using factors such as concepts, keywords, entities, number of friends and followers, website type, reputation, child safety, and website rating. The proposed framework was trained on six different classifiers, with the Support Vector Machine (SVM) achieving the best performance, scoring an accuracy of 86.30% and an F1 score of 86.33%. This research contributes to the development of systems capable of distinguishing credible news from non-credible content, thereby supporting users, researchers, and experts in ensuring the reliability of information shared online.