Early Detection of Fake News on Social Media Through Propagation Path Classification with Recurrent and Convolutional Networks

Abstract

In the midst of today's pervasive influence of social media, automatically detecting fake news is drawing significant attention from both the academic communities and the general public. Existing detection approaches rely on machine learning algorithms with a variety of news characteristics to detect fake news. However, such approaches have a major limitation on detecting fake news early, i.e., the information required for detecting fake news is often unavailable or inadequate at the early stage of news propagation. As a result, the accuracy of early detection of fake news is low. To address this limitation, in this paper, we propose a novel model for early detection of fake news on social media through classifying news propagation paths. We first model the propagation path of each news story as a multivariate time series in which each tuple is a numerical vector representing characteristics of a user who engaged in spreading the news. Then, we build a time series classifier that incorporates both recurrent and convolutional networks which capture the global and local variations of user characteristics along the propagation path respectively, to detect fake news. Experimental results on three real-world datasets demonstrate that our proposed model can detect fake news with accuracy 85% and 92% on Twitter and Sina Weibo respectively in 5 minutes after it starts to spread, which is significantly faster than state-of-the-art baselines.

DDGCN: Dual Dynamic Graph Convolutional Networks for Rumor Detection on Social Media

“…Ma et al (Ma et al 2016) exploits a recurrent neural network based model to capture the variation of semantics in propagation. Liu et al (Liu and Wu 2018) models the temporal structure by combining the recurrent and convolutional networks. Xia et al (Xia, Xuan, and Yu 2020) proposes a state-independent and time-evolving network for rumor detection based on fine-grained event state detection and segmentation.…”

Investigations of disaster information representation from a geospatial perspective: Progress, challenges and recommendations

“…Furthermore, to reduce the professionalism of disaster maps and make them more accessible to the public, the use of social media, such as WeChat and Open GIS, story maps, and open sources is advised. Moreover, panel discussions and questionnaires have been conducted in some studies to enable users to participate in the mapping process (Gaillard & Pangilinan, 2010; Liu & Wu, 2018; White, Kingston, & Barker, 2010). For example, a participatory GIS has been developed to engage the public in flood risk management and used participatory mapping to raise disaster risk awareness among youth in the Philippines.…”

“…For example, public awareness of Covid‐19 has been adversely affected by misinformation on social media. In terms of misinformation handling, a set of pandemic misinformation keywords (e.g., do not stay at home, city lockdown) can be developed, and then deep approaches, such as convolutional or recurrent neural network models, can be designed for automatically identifying and filtering misinformation from social media (Ajao, Bhowmik, & Zargari, 2018; Liu & Wu, 2018). Moreover, it is also meaningful to analyze the patterns and factors influencing misinformation dissemination from a geospatial analysis perspective (Forati & Ghose, 2021; Wang, Zhang, Fan, & Zhao, 2021).…”

“…Flood simulation software, such as MIKE (https://www.mikepoweredbydhi.com/), HEC (Hydrologic Engineering Center, https://www.hec.usace.army.mil/), FLO‐2D (https://flo-2d.com/), and Delft 3D (https://oss.deltares.nl/web/delft3d/), can be used to calculate the extent, flow velocity, and water depth. This information is overlaid with capacity/resource maps and social maps on GIS platforms to analyze the risk level of houses and roads, and then the final spatial distribution and accessibility are presented through map symbols and colors (Liu & Wu, 2018; Meyer et al, 2012; Ntajal, Lamptey, Mahamadou, & Nyarko, 2017; Symonds et al, 2016; Thakur, Parajuli, Kalra, Ahmad, & Gupta, 2017; Waldman et al, 2017; Wu, Liu, & Chen, 2013). However, cartographers and disaster management professionals guide the investigation and design of disaster maps and their thematic focus, which can often lead to the contents of disaster maps not matching the requirements of the public, and the representation methods cannot be easily understood (Holub & Fuchs, 2009; Meyer et al, 2012).…”

MetaDetector: Meta Event Knowledge Transfer for Fake News Detection

“…Social context-based detection methods principally embed social interactions (e.g., user comments/likes/reposts) or information difusion structures into dense vectors by neural networks for the following detection. For instance, Liu et al [22] observe that the true and fake news have diferent disseminated patterns, so they use the gated recurrent unit (GRU) and CNN to extract global and local features of the retweeting sequences for fake news detection. On this basis, Lu et al [26] introduce the graph convolution neural networks (GCNs) to learn more accurate structural information of news propagation paths.…”