Publication

Abstract : Social networking sites offer a vast platform for sharing various types of content, such as information, photos, videos, and audio clips. However, the credibility of the information shared on these sites is a significant concern since there is no prior verification process as the information spreads across the networks. This increases the chances of disseminating false or misleading information, often called rumors. The spread of rumors in today's information environment creates severe problems for the credibility of information and public confidence. The quick spread of rumors, frequently conveyed via social media and other online platforms, can affect public opinion and potentially have real-world repercussions. Maintaining the integrity of information ecosystems and averting the potentially damaging impacts of disinformation require detecting and mitigating rumors. In today's social media and online communication age, it is crucial to identify rumors to prevent their spread and mitigate their potential negative impact. In this study, our primary goal is to discover and identify individual sources of rumors. This work addresses two aspects of rumor management: firstly, identifying the difference between rumors and facts, and secondly, tracking the source of the rumor. In conjunction with the attributes, state-of-the-art methods such as SVM, AdaBoost, Random Forest, Logistic Regression, and Graph Convolutional Networks (GCN) are used to accomplish these goals. This approach contributes to the broader goal of combating disinformation and preserving the Accuracy of information.