Supervised Learning for Fake News Detection

Reis, Júlio C. S.; Correia, A. Gomes; Murai, Fabrício; Veloso, Adriano; Benevenuto, Fabrí­cio; Wang, Zhaoxia

doi:10.1109/mis.2019.2899143

Cited by 391 publications

(149 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…), emotional features, text topic features, etc. Subsequently, many linguistic-based approaches are developed based on deep learning (Reis et al, 2019;Ma et al, 2016;Wu et al, 2018;Ma et al, 2018b;Wu et al, 2019b). Typically, Ma et al (Ma et al, 2016) present a deep learning framework for rumor debunking, which learns RNN models by utilizing the variation of aggregated information across different time intervals related to each event.…”

Section: Information Credibility Evaluationmentioning

confidence: 99%

“…Deep semantics Ma et al 2016Ma et al 2018Wu et al 2018Reis et al 2019Rashkin et al 2017Martin et al 2018Tseng et al 1999Karimi et al 2018Long et al 2017Shu et al 2019 Profiles Influence Interests Yang et al 2018Ghenai et al 2018 Dynamic networks Static networks Jin et al 2016Tacchini et al 2017Ma et al 2017Ruchansky et al 2017Wu et al 2018 Stance-based Ma et al 2018Lukasik et al 2019 Meta-data based Figure 2. The review of information credibility evaluation methods Wu et al, 2019a).…”

Section: Shallow Semanticsmentioning

confidence: 99%

See 1 more Smart Citation

Discovering differential features: Adversarial learning for information credibility evaluation

Rao

Nazir

et al. 2020

Information Sciences

View full text Add to dashboard Cite

A series of deep learning approaches extract a large number of credibility features to detect fake news on the Internet. However, these extracted features still suffer from many irrelevant and noisy features that restrict severely the performance of the approaches. In this paper, we propose a novel model based on Adversarial Networks and inspirited by the Shared-Private model (ANSP), which aims at reducing common, irrelevant features from the extracted features for information credibility evaluation. Specifically, ANSP involves two tasks: one is to prevent the binary classification of true and false information for capturing common features relying on adversarial networks guided by reinforcement learning. Another extracts credibility features (henceforth, private features) from multiple types of credibility information and compares with the common features through two strategies, i.e., orthogonality constraints and KL-divergence for making the private features more differential. Experiments first on two six-label LIAR and Weibo datasets demonstrate that ANSP achieves the state-of-theart performance, boosting the accuracy by 2.1%, 3.1%, respectively and then on four-label Twit-ter16 validate the robustness of the model with 1.8% performance improvements.

show abstract

Section: Information Credibility Evaluationmentioning

confidence: 99%

Section: Shallow Semanticsmentioning

confidence: 99%

Discovering differential features: Adversarial learning for information credibility evaluation

Rao

Nazir

et al. 2020

Information Sciences

View full text Add to dashboard Cite

show abstract

“…Outros trabalhos [Ribeiro et al 2019, Andreou et al 2019, Speicher et al 2018 avaliaram o uso da plataforma de propagandas do Facebook para a realização de campanhas políticas que exploram o marketing direcionado como forma de disseminar propagandas falsas ou sobre temas divisíveis, incitando polarização é odio [Mondal et al 2017] nas redes sociais. Outros esforços visam explorar estratégias para detecção automática de notícias falsas [Reis et al 2019]. Nosso trabalhoé complementar aos esforços acima, pois investigamos a capacidade de viralização de um aplicativo que foi bastante explorado na disseminação de notícias falsas no Brasil.…”

Section: Trabalhos Relacionadosunclassified

O Paradoxo da Viralização de Informação Criptografada no WhatsApp

Vieira

Melo

et al. 2019

Anais Do XXXVII Simpósio Brasileiro De Redes De Computadores E Sistemas Distribuídos (SBRC 2019)

Self Cite

View full text Add to dashboard Cite

O WhatsApp é um sistema de comunicação móvel que permite que pessoas interajam através de grupos. Neste trabalho, analisamos adisseminacão de informações dentro de uma rede de grupos que simula a rede do WhatsApp. A rede construída considera dois tipos de grupos: grupos orgânicos, formados por amigos e familiares, e grupos artificiais que, em geral, são criados com o objetivo de ser um meio de divulgação de determinado assunto ou evento, como por exemplo, campanhas políticas. Analisamos a velocidade com que se dá o espalhamento de informação nessa rede considerando o modelo epidemiológico Suscetível-Infectado (SI). Em seguida aprofundamos nossa análise buscando identificar parâmetros que fazem com que esse espalhamento seja parcialmente controlado de forma a dificultar a propagação de notícias falsas nessas redes. Nossos resultados quantificam a capacidade de viralização de um conteúdo no WhatsApp e identificam aspectos que poderiam limitar tal capacidade para evitar que a plataforma seja abusada em períodos eleitorais.

show abstract

“…Reis et al [12] have worked on the dataset related to US election of year 2016 for the detection of fake news and posts on social media. The authors have tried to do the dataset based experimentation along with the real time practice experimentation.…”

Section: Literature Reviewmentioning

confidence: 99%

Ant Colony Decision Tree Method to Detect the Suspicious News

Kumari*,

Balkishan

2019

IJRTE

View full text Add to dashboard Cite

The ease of availability and low cost of social media platforms has made the human vulnerable to devour the information from social media. Social media serves the unverified information which easily gets disseminated among people through different groups, applications, and other online social platforms. These fake news can lead to any suspicious activities which urges to derive the term suspicious news. The escalation of these suspicious news has open a lot of research ventures to detect and attenuate the impact on human. These suspicious activities have become a nuisance for legitimate users. Despite the presence of existing methods for the suspicious news detection exists, but the continuous growth of such activities is difficult to manage with an individual approach. In this research work, an ensemble approach is considered to detect suspicious news content. Here, Ant Colony Decision Tree method is ensembled for the detection of suspicious news (ACDTDSN). This approach uses the heuristic function and pheromone trail to obtain the optimal solution. The overall functionality of system is based on the content based approach for the detection of suspicious news with steps of dataset consideration, pre-processing, feature selection, and classification. The experimentation is performed using the FakenewNet dataset which consist of BuzzFeed and PolitiFact categories of news content. The results of the proposed ACSTDSN framework are accessed with the performance evaluation measures.

show abstract

Supervised Learning for Fake News Detection

Cited by 391 publications

References 9 publications

Discovering differential features: Adversarial learning for information credibility evaluation

Discovering differential features: Adversarial learning for information credibility evaluation

O Paradoxo da Viralização de Informação Criptografada no WhatsApp

Ant Colony Decision Tree Method to Detect the Suspicious News

Contact Info

Product

Resources

About