Fake news believability: The effects of political beliefs and espoused cultural values

Fake news has led to a polarized society as evidenced by diametrically opposed perceptions of and reactions to global events such as the Coronavirus Disease 2019 (COVID-19) pandemic and presidential campaigns. Popular press has linked individuals’ political beliefs and cultural values to the extent to which they believe in false content shared on social networking sites (SNS). However, sweeping generalizations run the risk of helping exacerbate divisiveness in already polarized societies. This study examines the effects of individuals’ political beliefs and espoused cultural values on fake news believability using a repeated-measures design (that exposes individuals to a variety of fake news scenarios). Results from online questionnaire-based survey data collected from participants in the US and India help confirm that conservative individuals tend to exhibit increasing fake news believability and show that collectivists tend to do the same. This study advances knowledge on characteristics that make individuals more susceptible to lending credence to fake news. In addition, this study explores the influence exerted by control variables (i.e., age, sex, and Internet usage). Findings are used to provide implications for theory as well as actionable insights.     

基于区块链技术的虚假新闻检测方法

虚假新闻不仅会导致人们形成错误观念,损害人们的知情权,还会降低新闻网站公信力。针对新闻网站出现虚假新闻的问题,提出一种基于区块链技术的虚假新闻检测方法。首先,通过调用智能合约为新闻随机分配审核者来判定新闻的真实性。然后,调整审核者数量以确保有效审核者的数量,提高审核结果的可信度。同时设计激励机制,根据审核者的行为分配奖励,并运用博弈论分析审核者的行为和获得的奖励,为了获得最大利益,审核者的行为必须是诚实的。而后设计审计机制检测恶意的审核者,以提高系统的安全性。最后,利用以太坊智能合约实现了一个简易的区块链虚假新闻检测系统,并对虚假新闻检测进行了仿真,结果显示所提方法的新闻真实性检测的准确率达到了95%,表明该方法可有效防止虚假新闻的发布。     

Fake News Identification and Classification Using DSSM and Improved Recurrent Neural Network Classifier

ABSTRACT

The widespread use of social media has enormous consequences for the society, culture and business with potentially positive and negative effects. As online social networks are increasingly used for dissemination of information, at the same they are also becoming a medium for the spread of fake news for various commercial and political purposes. Technologies such as Artificial Intelligence (AI) and Natural Language Processing (NLP) tools offer great promise for researchers to build systems, which could automatically detect fake news. However, detecting fake news is a challenging task to accomplish as it requires models to summarize the news and compare it to the actual news in order to classify it as fake. This project proposes a framework that detects and classifies fake news messages using improved Recurrent Neural Networks and Deep Structured Semantic Model. The proposed approach intuitively identifies important features associated with fake news without previous domain knowledge while achieving accuracy 99%. The performance analysis method used for the proposed system is based on accuracy, specificity and sensitivity.     

基于多模双线性池化方法的虚假新闻检测模型

社交媒体的兴起促进了新闻行业的发展,使虚假新闻的传播也变得更为便利,然而多样化的新闻表现形式带来了很多负面影响,比如新闻内容夸大事实、恶意篡改新闻文本或图像内容、构造虚假新闻事实引起社会舆论,这促使了虚假新闻检测工作成为新闻领域新的挑战。为了应对虚假新闻检测工作的研究,将新闻文本与图像信息结合起来,通过多模双线性池化方法,改变传统特征融合方法,构建出基于新特征融合方法的虚假新闻检测模型,并且采用虚假新闻检测领域标准数据集验证模型的性能,实验结果表明,文本与图像的融合特征表现在虚假新闻检测领域不可替代,且所提方法能够有效提升虚假新闻检测性能。     

Integrating truth bias and elaboration likelihood to understand how political polarisation impacts disinformation engagement on social media

Political polarisation has become an increasingly alarming issue in society, exacerbated by the widespread use of social media and the development of filter bubbles among social media users. This environment has left users susceptible to disinformation, especially those with whom a user is politically aligned. In this research, we integrate truth bias, elaboration likelihood model and new media literacy into a model for explaining social media engagement (with both disinformation and factual information) and analysing how political polarisation (operationalised as political alignment between users) influences perceptions and behaviours. Using an experimental design, we analyse the model separately for posts containing disinformation and factual information, highlighting key differences. Political alignment positively moderates truth bias's effect on engagement with disinformation. For both disinformation and factual information, political alignment moderates the effect of generalised communicative suspicion (GCS) on truth bias, such that GCS's effect on truth bias flips from negative to positive as political alignment increases. Issue involvement and political alignment appear to be the primary drivers of disinformation engagement, with critical consuming media literacy failing to mitigate engagement. Our findings contribute to the understanding of persuasion, conviction, amplification, polarisation and aversion related to fake news on social media.     

Distance-based customer detection in fake follower markets

As the Online Social Networks (OSNs) have become popular, more and more people want to increase their influence not only in the real world but also in the OSNs. However, increasing the influence in OSNs is time-consuming job, so some users want to find a shortcut to increase their relationships. The demand for quick increasement of relationship has led to the growth of the fake follower markets that cater to customers who want to grow their relationships rapidly. However, customers of fake follower markets cannot manipulate legitimate user’s relationship perfectly. Existing approaches explore node’s relationships or features to detect customers. But none of them combines the relationships and node’s features directly. In this article, we propose a model that directly combines the relationship and node’s feature to detect customers of fake followers. Specifically, we study the geographical distance for 1-hop-directional links using the nodes geographical location. Motivated by the difference of a distance ratio for 1-hop directional links, the proposed method is designed to generate a 1-hop link distance ratio, and classifies a node as a customer or not. Experimental results on a Twitter dataset demonstrate that the proposed method achieves higher performance than other baseline methods.     

Rumour detection using deep learning and filter-wrapper feature selection in benchmark twitter dataset

Multimedia Tools and Applications - Microblogs have become a customary news media source in recent times. But as synthetic text or ‘readfakes’ scale up the online disinformation...     

DTN: Deep triple network for topic specific fake news detection

Detection of fake news has spurred widespread interests in areas such as healthcare and Internet societies, in order to prevent propagating misleading information for commercial and political purposes. However, efforts to study a general framework for exploiting knowledge, for judging the trustworthiness of given news based on their content, have been limited. Indeed, the existing works rarely consider incorporating knowledge graphs (KGs), which could provide rich structured knowledge for better language understanding.In this work, we propose a deep triple network (DTN) that leverages knowledge graphs to facilitate fake news detection with triple-enhanced explanations. In the DTN, background knowledge graphs, such as open knowledge graphs and extracted graphs from news bases, are applied for both low-level and high-level feature extraction to classify the input news article and provide explanations for the classification.The performance of the proposed method is evaluated by demonstrating abundant convincing comparative experiments. Obtained results show that DTN outperforms conventional fake news detection methods from different aspects, including the provision of factual evidence supporting the decision of fake news detection.     

Social interaction for efficient agent learning from human reward

Learning from rewards generated by a human trainer observing an agent in action has been proven to be a powerful method for teaching autonomous agents to perform challenging tasks, especially for those non-technical users. Since the efficacy of this approach depends critically on the reward the trainer provides, we consider how the interaction between the trainer and the agent should be designed so as to increase the efficiency of the training process. This article investigates the influence of the agent’s socio-competitive feedback on the human trainer’s training behavior and the agent’s learning. The results of our user study with 85 participants suggest that the agent’s passive socio-competitive feedback—showing performance and score of agents trained by trainers in a leaderboard—substantially increases the engagement of the participants in the game task and improves the agents’ performance, even though the participants do not directly play the game but instead train the agent to do so. Moreover, making this feedback active—sending the trainer her agent’s performance relative to others—further induces more participants to train agents longer and improves the agent’s learning. Our further analysis shows that agents trained by trainers affected by both the passive and active social feedback could obtain a higher performance under a score mechanism that could be optimized from the trainer’s perspective and the agent’s additional active social feedback can keep participants to further train agents to learn policies that can obtain a higher performance under such a score mechanism.     

User attitudes towards news content personalization

Personalizing news content requires to choose the appropriate depth of personalization and to assess the extent to which readers’ explicit expressions of interest in general and specific news topics can be used as the basis for personalization. A preliminary survey examined 117 respondents’ attitudes towards news content personalization and their interest in various news topics and subtopics. The second survey examined 23 participants’ declared and actual interests. Participants preferred personalization based on general news topics. Declared interest in general news topics adequately predicted the actual interests in some topics, while in others users’ interests differed between general news topics and subtopics. The variance in interest in items also differed among topics. Thus, different personalization methods should be used for different topics. For some, such as ‘Sports’, users show either high interest or no interest at all. In the latter case most articles related to the topic should be removed, with the exception of items that refer to unique events that may raise general interest according to the expressed interest. In other topics, such as ‘Science & Technology’, most users are interested in important articles, even if they are not interested in the general news topic. Here, the filtering technique should identify the important articles and present them to all readers. The results can be used to develop effective and simple personalization mechanisms which can be applied to the personalization of news, as well as to other domains.     

网络传播信息内容的可信度研究进展

网络中存在着大量的谣言、偏激和虚假信息,这对网络信息的质量、可信度以及舆情的产生与发展趋势具有严重的负面影响。为实现信息可信度的准确判断与高效度量,该文在大量已有最新研究成果与文献的基础上,将不可信信息分为极端突发事件信息、网络偏激信息、网络谣言、虚假信息、误报信息和垃圾信息等类型,并分别针对这些类型信息从分类定义、内容特征描述、可信度建模以及可信度评测等四个方面进行研究综述,从而为网络传播中信息内容的可信度分析与度量研究奠定坚实基础。最后,进一步对信息可信度研究的发展方向进行展望。     

Using think-aloud and psychometrics to explore users’ experience with a news Web site

The present study is part of a research programme that aims to develop and test a psychological model of end-users’ experience with news sites. An exploratory study of interaction experience with a news Web site was conducted. An online questionnaire was used to collect information on demographics, Internet-use and news-site use behaviour of users of a particular news site, and to recruit participants for a think-aloud study. The protocol analysis of screen-capture and audio recordings of participants, who used a news site while thinking aloud, yielded five categories of experience: impression, content, layout, information architecture and diversion. These categories are regarded as spontaneous, self-reported aspects of users’ experience with a news site. A set of interaction-experience questionnaires revealed significant differences between regular users and non-users of a news site. Correlation and regression analyses demonstrated support for Hassenzahl’s model of interaction experience. The study presents a first attempt to empirically investigate the aspects of interaction experience in relation to online news sites.     

An effective strategy for multi-modal fake news detection

News plays an indispensable role in the development of human society. With the emergence of new media, fake news including multi-modal content such as text and images has greater social harm. Therefore how to identify multi-modal fake news has been a challenge. The traditional methods of multi-modal fake news detection are to simply fuse the different modality information, such as concatenation and element-wise product, without considering the different impacts of the different modalities, which leads to the low accuracy of fake news detection. To address this issue, we design a new multi-modal attention adversarial fusion method built on the pre-training language model BERT, which consists of two important components: the attention mechanism and the adversarial mechanism. The attention mechanism is used to capture the differences in different modalities. The adversarial mechanism is to capture the correlation between different modalities. Experiments on a fake news Chinese public dataset indicate that our proposed new method achieves 5% higher in terms of F1.

     

HeadlineStanceChecker: Exploiting summarization to detect headline disinformation

The headline of a news article is designed to succinctly summarize its content, providing the reader with a clear understanding of the news item. Unfortunately, in the post-truth era, headlines are more focused on attracting the reader’s attention for ideological or commercial reasons, thus leading to mis- or disinformation through false or distorted headlines. One way of combating this, although a challenging task, is by determining the relation between the headline and the body text to establish the stance. Hence, to contribute to the detection of mis- and disinformation, this paper proposes an approach (HeadlineStanceChecker) that determines the stance of a headline with respect to the body text to which it is associated. The novelty rests on the use of a two-stage classification architecture that uses summarization techniques to shape the input for both classifiers instead of directly passing the full news body text, thereby reducing the amount of information to be processed while keeping important information. Specifically, summarization is done through Positional Language Models leveraging on semantic resources to identify salient information in the body text that is then compared to its corresponding headline. The results obtained show that our approach achieves 94.31% accuracy for the overall classification and the best FNC-1 relative score compared with the state of the art. It is especially remarkable that the system, which uses only the relevant information provided by the automatic summaries instead of the whole text, is able to classify the different stance categories with very competitive results, especially in the discuss stance between the headline and the news body text. It can be concluded that using automatic extractive summaries as input of our approach together with the two-stage architecture is an appropriate solution to the problem.     

The mass,fake news,and cognition security

The widespread fake news in social networks is posing threats to social stability, economic development, and political democracy, etc. Numerous studies have explored the effective detection approaches of online fake news, while few works study the intrinsic propagation and cognition mechanisms of fake news. Since the development of cognitive science paves a promising way for the prevention of fake news, we present a new research area called Cognition Security (CogSec), which studies the potential impacts of fake news on human cognition, ranging from misperception, untrusted knowledge acquisition, targeted opinion/attitude formation, to biased decision making, and investigates the effective ways for fake news debunking. CogSec is a multidisciplinary research field that leverages the knowledge from social science, psychology, cognition science, neuroscience, AI and computer science. We first propose related definitions to characterize CogSec and review the literature history. We further investigate the key research challenges and techniques of CogSec, including humancontent cognition mechanism, social influence and opinion diffusion, fake news detection, and malicious bot detection. Finally, we summarize the open issues and future research directions, such as the cognition mechanism of fake news, influence maximization of fact-checking information, early detection of fake news, fast refutation of fake news, and so on.     

A neurophysiological approach to assess training outcome under stress: A virtual reality experiment of industrial shutdown maintenance using Functional Near-Infrared Spectroscopy (fNIRS)

Shutdown maintenance, i.e., turning off a facility for a short period for renewal or replacement operations is a highly stressful task. With the limited time and complex operation procedures, human stress is a leading risk. Especially shutdown maintenance workers often need to go through excessive and stressful on-site trainings to digest complex operation information in limited time. The challenge is that workers’ stress status and task performance are hard to predict, as most trainings are only assessed after the shutdown maintenance operation is finished. A proactive assessment or intervention is needed to evaluate workers’ stress status and task performance during the training to enable early warning and interventions. This study proposes a neurophysiological approach to assess workers’ stress status and task performance under different virtual training scenarios. A Virtual Reality (VR) system integrated with the eye-tracking function was developed to simulate the power plant shutdown maintenance operations of replacing a heat exchanger in both normal and stressful scenarios. Meanwhile, a portable neuroimaging device – Functional Near-Infrared Spectroscopy (fNIRS) was also utilized to collect user’s brain activities by measuring hemodynamic responses associated with neuron behavior. A human–subject experiment (n = 16) was conducted to evaluate participants’ neural activity patterns and physiological metrics (gaze movement) related to their stress status and final task performance. Each participant was required to review the operational instructions for a pipe maintenance task for a short period and then perform the task based on their memory in both normal and stressful scenarios. Our experiment results indicated that stressful training had a strong impact on participants’ neural connectivity patterns and final performance, suggesting the use of stressors during training to be an important and useful control factors. We further found significant correlations between gaze movement patterns in review phase and final task performance, and between the neural features and final task performance. In summary, we proposed a variety of supervised machine learning classification models that use the fNIRS data in the review session to estimate individual’s task performance. The classification models were validated with the k-fold (k = 10) cross-validation method. The Random Forest classification model achieved the best average classification accuracy (80.38%) in classifying participants’ task performance compared to other classification models. The contribution of our study is to help establish the knowledge and methodological basis for an early warning and estimating system of the final task performance based on the neurophysiological measures during the training for industrial operations. These findings are expected to provide more evidence about an early performance warning and prediction system based on a hybrid neurophysiological measure method, inspiring the design of a cognition-driven personalized training system for industrial workers.     

Optimal Weighted Extreme Learning Machine for Cybersecurity Fake News Classification

Fake news and its significance carried the significance of affecting diverse aspects of diverse entities, ranging from a city lifestyle to a country global relativity, various methods are available to collect and determine fake news. The recently developed machine learning (ML) models can be employed for the detection and classification of fake news. This study designs a novel Chaotic Ant Swarm with Weighted Extreme Learning Machine (CAS-WELM) for Cybersecurity Fake News Detection and Classification. The goal of the CAS-WELM technique is to discriminate news into fake and real. The CAS-WELM technique initially pre-processes the input data and Glove technique is used for word embedding process. Then, N-gram based feature extraction technique is derived to generate feature vectors. Lastly, WELM model is applied for the detection and classification of fake news, in which the weight value of the WELM model can be optimally adjusted by the use of CAS algorithm. The performance validation of the CAS-WELM technique is carried out using the benchmark dataset and the results are inspected under several dimensions. The experimental results reported the enhanced outcomes of the CAS-WELM technique over the recent approaches.     

Optimal Quad Channel Long Short-Term Memory Based Fake News Classification on English Corpus

The term ‘corpus’ refers to a huge volume of structured datasets containing machine-readable texts. Such texts are generated in a natural communicative setting. The explosion of social media permitted individuals to spread data with minimal examination and filters freely. Due to this, the old problem of fake news has resurfaced. It has become an important concern due to its negative impact on the community. To manage the spread of fake news, automatic recognition approaches have been investigated earlier using Artificial Intelligence (AI) and Machine Learning (ML) techniques. To perform the medicinal text classification tasks, the ML approaches were applied, and they performed quite effectively. Still, a huge effort is required from the human side to generate the labelled training data. The recent progress of the Deep Learning (DL) methods seems to be a promising solution to tackle difficult types of Natural Language Processing (NLP) tasks, especially fake news detection. To unlock social media data, an automatic text classifier is highly helpful in the domain of NLP. The current research article focuses on the design of the Optimal Quad Channel Hybrid Long Short-Term Memory-based Fake News Classification (QCLSTM-FNC) approach. The presented QCLSTM-FNC approach aims to identify and differentiate fake news from actual news. To attain this, the proposed QCLSTM-FNC approach follows two methods such as the pre-processing data method and the Glove-based word embedding process. Besides, the QCLSTM model is utilized for classification. To boost the classification results of the QCLSTM model, a Quasi-Oppositional Sandpiper Optimization (QOSPO) algorithm is utilized to fine-tune the hyperparameters. The proposed QCLSTM-FNC approach was experimentally validated against a benchmark dataset. The QCLSTM-FNC approach successfully outperformed all other existing DL models under different measures.     

基于矩阵分解双线性池化的多模态融合虚假新闻检测

现有的大多数虚假新闻检测方法将视觉和文本特征串联拼接,导致模态信息冗余并且忽略了不同模态信息之间的相关性。为了解决上述问题,提出一种基于矩阵分解双线性池化的多模态融合虚假新闻检测算法。首先,该算法将多模态特征提取器捕捉的文本和视觉特征利用矩阵分解双线性池化方法进行有效融合,然后与虚假新闻检测器合作鉴别虚假新闻;此外,在训练阶段加入了事件分类器来预测事件标签并去除事件相关的依赖。在Twitter和微博两个多模态谣言数据集上进行了对比实验,证明了该算法的有效性。实验结果表明提出的模型能够有效地融合多模态数据,缩小模态间的异质性差异,从而提高虚假新闻检测的准确性。     

一种基于双流网络的Deepfakes检测技术

随着深度学习技术的飞速发展,以Deepfakes为代表的深度伪造技术开始充斥在互联网上的各个角落。Deepfakes借助于生成对抗网络和自动编码器技术,能够轻松替换人脸以及篡改人的表情信息。此类Deepfakes假视频可以制作虚假色情影片、谣言,传播假新闻,甚至影响政治选举,带来的社会影响极其恶劣。然而,针对此类伪造视频的检测技术还远远落后于生成技术,已有的工作都存在一定的局限性,并不能较好地对Deepfakes视频进行检测。本文首先对现有生成和检测工作进行综述,并分析了现有工作的缺陷,然后提出了基于EfficientNet的双流网络检测框架。通过在大规模开源数据集FaceForensics++测试,我们的检测技术可以在检测Deepfakes类假视频上平均准确率达到99%以上,并一定程度提高模型对抗压缩的能力。