@misc{17666, author = {Asep Maulana and Johannes Langguth}, title = {Using GNNs for Misinformation Spreader Detection via Assortativity-Aware Node Label Classification in Twitter Networks}, abstract = {Misinformation on social media platforms such as Twitter and Facebook has become more prevalent and poses a significant challenge for ensuring the integrity of information ecosystems. Identifying key actors responsible for spreading misinformation is crucial in developing effective strategies to mitigate its impact. This paper presents an analysis of network structure and probability analysis of misinformation spreaders in the Twitter network by applying matrix correlation, attribute assortativity and Graph Neural Networks (GNNs) for node label classification. Attribute assortativity is relevant in understanding connected nodes and engagement patterns. In this study, we employ GNNs, a powerful machine learning technique for node classification based on node features in network data, using human-generated labels as the ground truth. By leveraging the structural information of the network, GNNs capture the complex relationships and propagation patterns among nodes, allowing for an accurate classification of misinformation spreaders. To evaluate the proposed methodology, a Twitter dataset of network data, user attributes and information propagation patterns, is utilized. The dataset is derived from the the Twitter network, enhanced with ground truth labels in nine different categories of misinformation topics indicating whether a user is a misinformation spreader or not. Despite the lack of a clear effect of different attribute assortativity among various categories on classification performance, our research has unequivocally confirmed that, within the category demonstrating the highest accuracy in GNNs classification, nodes affiliated with the misinformation spreader class display the highest likelihood of propagating misinformation.}, year = {2023}, journal = {2023 Tenth International Conference on Social Networks Analysis, Management and Security (SNAMS)}, publisher = {IEEE}, url = {https://ieeexplore.ieee.org/abstract/document/10375407}, }