Fake News Research Articles

Disinformation and political propaganda: An exploration of the risks of artificial intelligence

A significant shift is currently underway in the disinformation industry. We are transitioning from the era of disinformation fuelled by fake news and social media to disinformation on a larger scale generated through artificial intelligence (AI). Therefore, the objective of this text is to analyse this disinformation phenomenon, catalysed by social media and AI, from the media ecology perspective. This work is divided into two parts. In the first part of the text, we analyse the disinformation phenomenon, highlighting the involvement of certain governments. In the second part of the text, we focus on recognizing the effects that can arise from the use of AI within the extensive landscape of the disinformation industry.

The digital revolution contributed to practical reality according to the stages of its emergence in contemporary history (A model)

The digital revolution has politically affected the actual political reality through the transformation of the relatively free digital public sphere, under official digital censorship systems, and also for major global digital companies. Digitalization has today become a tool for building digital pressure groups that exercise their criticism and demands outside partisan frameworks and outside constitutional and legal systems, and outside The traditional mechanisms of a representative democratic system, including party pluralism, competitions, parliamentary and party elections within each party, municipal elections, referendums, and the formation of governments based on the results of the elections, or presidential elections in presidential systems such as the United States of America, or parliamentary and semi-presidential elections in the French style. Therefore, this study seeks to highlight knowledge of several questions about the possibilities of modifying the system of representative democracy towards introducing some forms of direct democracy, and not waiting for general elections to change the policies of elected governments, according to the political phenomena of digitization and its impact on democratic political culture as part of the processes of liquidity, and transformation in terms of the emergence of some phenomena. Populism, the post-truth world, and fake news in politics.

Values in British higher education: knowledge, freedom and wellbeing

British higher education is ranked among the best in the world, but some of its core values are under strain. Knowledge and critical thinking can be undermined by biased mind-sets, especially when engaging with the social media. Research demonstrates that false news goes viral much more quickly than true news. Political correctness and the woke movement can militate against freedom, especially in gender-related matters. Surveys show that many staff and students have a lower sense of wellbeing than the rest of the population. Conscious of these problems, the Government has responded with an Act to place extensive new obligations on certain types of online service providers, requiring them to protect their users by managing risks relating to illegal and harmful content. A separate Act has also been passed to protect freedom of speech in universities. The issue of staff and student unhappiness is much less amenable to legislation because it is so multi-faceted. However, it is a very important issue for a healthy, democratic society that fosters cooperation, trust and community. It needs to be addressed.

Criminal liability arising from spreading rumors and false news via social media sites

Criminal liability arising from spreading rumors and false news via social media sites

Graph contrastive learning for source localization in social networks

In real-world scenarios, rapidly and accurately identifying sources of fake news or disease outbreaks is crucial for public safety. Existing deep learning methods for source localization tasks entirely rely on supervised learning, which requires a large amount of labeled data. Recently, the emerging self-supervised learning (SSL) methods have significantly reduced the reliance on labeled data. Nevertheless, the investigation of SSL for source localization tasks remains unexplored. In this work, we are the first to adapt SSL for source localization tasks, specifically employing graph contrastive learning (GCL). Yet, directly applying GCL to source localization tasks faces two challenges: 1) existing data augmentation strategies are not well-suited for source localization tasks; 2) extremely low-dimensional node features potentially compromise the quality of learned node representations. To address these challenges, we introduce a Source Localization with Graph Contrastive Learning (SL-GCL) framework. Firstly, we propose a data augmentation strategy which exploits the inherent stochasticity of propagation. Secondly, we design a feature enrichment module to expand the feature dimensions. Finally, our experiments on six real-world networks demonstrate that SL-GCL outperforms state-of-the-art methods and exhibits remarkable transferability across different networks and propagation patterns.

Fake news detection models using the largest social media ground-truth dataset (TruthSeeker)

Fake news detection models using the largest social media ground-truth dataset (TruthSeeker)

Frontier Fake News: Nevada's Sagebrush Humorists and Hoaxsters by Richard Moreno (review)

Frontier Fake News: Nevada's Sagebrush Humorists and Hoaxsters by Richard Moreno (review)

Evaluating text classification: A benchmark study

This paper presents an impartial and extensive benchmark for text classification involving five different text classification tasks, 20 datasets, 11 different model architectures, and 42,800 algorithm runs. The five text classification tasks are fake news classification, topic detection, emotion detection, polarity detection, and sarcasm detection. While in practice, especially in Natural Language Processing (NLP), research tends to focus on the most sophisticated models, we hypothesize that this is not always necessary. Therefore, our main objective is to investigate whether the largest state-of-the-art (SOTA) models are always preferred, or in what cases simple methods can compete with complex models, i.e. for which dataset specifications and classification tasks. We assess the performance of different methods with varying complexity, ranging from simple statistical and machine learning methods to pretrained transformers like robustly optimized BERT (Bidirectional Encoder Representations from Transformers) pretraining approach (RoBERTa). This comprehensive benchmark is lacking in existing literature, with research mainly comparing similar types of methods. Furthermore, with increasing awareness of the ecological impacts of extensive computational resource usage, this comparison is both critical and timely.We find that overall, bidirectional long short-term memory (LSTM) networks are ranked as the best-performing method albeit not statistically significantly better than logistic regression and RoBERTa. Overall, we cannot conclude that simple methods perform worse although this depends mainly on the classification task. Concretely, we find that for fake news classification and topic detection, simple techniques are the best-ranked models and consequently, it is not necessary to train complicated neural network architectures for these classification tasks. Moreover, we also find a negative correlation between F1 performance and complexity for the smallest datasets (with dataset size less than 10,000). Finally, the different models’ results are analyzed in depth to explain the model decisions, which is an increasing requirement in the field of text classification.

Examining the growth in willingness to pay for digital wellbeing services on social media: A comparative analysis

In recent years, there has been a growing need for social media platforms to offer services that preserve and promote users' digital wellbeing, including better protection of personal data and balanced technology usage. However, the current business model of social media is often seen as in conflict with users' digital wellbeing. In 2020, a study investigated users' willingness to pay monetary fees for social media digital wellbeing services. In the present work, we replicated this study in Q4 of 2022, aiming to explore any changes in interest and willingness to pay for these services. In addition, we extended the replication by conducting qualitative analysis on participants' comments to gain deeper insight and identify reasons for payment and reasons for rejecting to pay. Data were collected from 262 participants through an online questionnaire. The survey focused on four services: better data protection, less use of data for marketing, aiding users in controlling their prolonged usage, and reducing fake news and radicalisation on social media. The results showed that the willingness to pay for these services was significantly higher in 2022 compared to the results published in 2020. Participants expressed concerns about the feasibility and fairness of the alternative business model, which requires users to pay for safety and support. Our findings suggest a growing interest in digital wellbeing services, emphasizing the need for social media platforms to assess the feasibility of alternative business models, identify user segments, and take measures to enhance consumers’ trust, accordingly.

Towards Optimal NLP Solutions: Analyzing GPT and LLaMA-2 Models Across Model Scale, Dataset Size, and Task Diversity

This study carries out a comprehensive comparison of fine-tuned GPT models (GPT-2, GPT-3, GPT-3.5) and LLaMA-2 models (LLaMA-2 7B, LLaMA-2 13B, LLaMA-2 70B) in text classification, addressing dataset sizes, model scales, and task diversity. Since its inception in 2018, the GPT series has been pivotal in advancing NLP, with each iteration introducing substantial enhancements. Despite its progress, detailed analyses, especially against competitive open-source models like the LLaMA-2 series in text classification, remain scarce. The current study fills this gap by fine-tuning these models across varied datasets, focusing on enhancing task-specific performance in hate speech and offensive language detection, fake news classification, and sentiment analysis. The learning efficacy and efficiency of the GPT and LLaMA-2 models were evaluated, providing a nuanced guide to choosing optimal models for NLP tasks based on architectural benefits and adaptation efficiency with limited data and resources. In particular, even with datasets as small as 1,000 rows per class, the F1 scores for the GPT-3.5 and LLaMA-2 models exceeded 0.9, reaching 0.99 with complete datasets. Additionally, the LLaMA-2 13B and 70B models outperformed GPT-3, demonstrating their superior efficiency and effectiveness in text classification. Both the GPT and LLaMA-2 series showed commendable performance on all three tasks, underscoring their ability to handle a diversity of tasks. Based on the size, performance, and resources required for fine-tuning the model, this study identifies LLaMA-2 13B as the most optimal model for NLP tasks.

A Federated Convolution Transformer for Fake News Detection

A Federated Convolution Transformer for Fake News Detection

Report on the 4th Workshop on Reducing Online Misinformation through Credible Information Retrieval (ROMCIR 2024) at ECIR 2024

The 4th Workshop on Reducing Online Misinformation through Credible Information Retrieval (ROMCIR 2024) was part of the Satellite Events of the 46th European Conference on Information Retrieval (ECIR 2024). ROMCIR served as a platform for discussions on accessing accurate information and addressing the issue of information disorder prevalent in the online landscape at that time. In general, the challenge is multifaceted, encompassing various types of information sources (e.g., websites, social media posts, ...) across different platforms and domains (e.g., fake news detection, health-related information retrieval, propaganda reduction). Additionally, in this edition, the critical need to assess the impact of generative models like Large Language Models (LLMs) on inadvertently amplifying misinformation and explore their potential role in supporting Information Retrieval Systems (IRSs) began to emerge. In this context, diverse approaches to the problem of access to truthful information found their place. Keynote speech and articles in this year's workshop mainly focused on themes such as health misinformation, multimedia and multimodal fact-checking, and information filtering to combat misinformation. Date : 24 March 2024. Website : https://romcir.disco.unimib.it/.

Disinformation and "hate speech" in the online media of Kazakhstan (results of monitoring 2022)

The article examines the problem of using the "language of hostility" and disinformation in the media texts of the Kazakh media, as well as the problem of stuffing fake news during the political events of January 2022 in Kazakhstan. The goals of disinformation are determined and the effects on the mass audience are evaluated. During the tragic events at the beginning of the year, the Internet was turned off in the country, which led to an information vacuum. After the restoration of communication, a stream of fake materials of various formats (audio, video, images and messages on social networks) began, the purpose of which was to incite fears among the population. The platforms WhatsApp, YouTube, Telegram and VKontakte were investigated. Case analysis of texts was also used as part of the analysis. The materials were selected by random sampling. Various criteria were taken into account for the selection of publications, including attendance, the toxicity of comments, the number of articles and posts in online publications and social networks, as well as genres of materials and the presence of elements of hate speech. The purpose of this study is to comprehensively analyze the destructive consequences of disinformation recorded in the Kazakh and Russian media during the attempt to forcibly seize power in the Republic of Kazakhstan. The following methods were used: critical analysis of scientific literature, media monitoring, discourse analysis, and the method of comparative analysis. Case analysis of texts was also used. In 2022, Kazakhstan faced new geopolitical and information challenges: armed border conflicts, disinformation and "hate speech". All these risks have led to an even greater polarization of Kazakh society, socio-political tension and a decrease in the level of trust in the official media. The key factor in changing the situation may be increasing the information literacy of the population and its promotion through cooperation with civil society institutions, the implementation of universal media education programs. Given the volume of data and the amount of information processed, it is difficult to fit the whole situation and analyze how the discourse of "hate speech" and disinformation was built. The repeatedly disseminated unreliable and contradictory information significantly increased the tension of the network media field, contributing to a greater polarization of opinions about the extraordinary event and the aggression of participants in the information exchange.

Hybrid model: IndoBERT and long short-term memory for detecting Indonesian hoax news

The world has entered an era that technology has developed far. Due to rapid technological development, information is easily spread. However, not all information spread through social media is factual information. Responding to this social phenomenon, we initiated to create a hoax detection system using the combined method of Indo bidirectional encoder representations from transformers (IndoBERT) and long short-term memory (LSTM). The dataset used in this study are obtained through the process scraping on the site turnbackhoax.id and cable news network (CNN) Indonesia. We decided to use the IndoBERT-LSTM method to detect hoaxes, using IndoBERT as the feature extractor and LSTM as the classification layer can be an effective method because of its advantages in managing and understanding Indonesian language. The results show that the IndoBERT-LSTM model achieved an accuracy of 93.2%, precision of 92%, recall of 89.7%, and F1-score of 90,8%. From a total of 5876 data composed of a total of 1998 factual news and 3878 hoax data. The hoax detection system using IndoBERT-LSTM is a promising approach for detecting hoaxes accurately and efficiently. This model has the potential to make a significant impact in the fight against the spread of Hoaxes.

A Multi-tier Model and Filtering Approach to Detect Fake News Using Machine Learning Algorithms

Fake news trends have overgrown in our societies over the years through social media platforms. The goal of spreading fake news can easily mislead and manipulate the public’s opinion. Many previous researchers have proposed this domain using classification algorithms or deep learning techniques. However, machine learning algorithms still suffer from high margin error, which makes them unreliable as every algorithm uses a different way of prediction. Deep learning requires high computation power and a large dataset to operate the classification model. A filtering model with a consensus layer in a multi-tier model is introduced in this research paper. The multi-tier model filters the news label correctly predicted by the first two-tier layer. The consensus layer acts as the final decision when collision results occur in the first two-tier layer. The proposed model is applied to the WEKA software tool to test and evaluate the model from both datasets. Two sequences of classification models are used in this research paper: LR_DT_RF and LR_NB_AdaBoost. The best performance of sequence for both datasets is LR_DT_RF which yields 0.9892 F1-Score, 0.9895 Accuracy, and 0.9790 Matthews Correlation Coefficient (MCC) for ISOT Fake News Dataset, and 0.9913 F1-Score, 0.9853 Accuracy, and 0.9455 MCC for CHECKED Dataset. This research could give researchers an approach for fake news detection on different social platforms and feature-based

Combating Misinformation: Leveraging Deep Learning for Hoax Detection in Indonesian Political Social Media

The rampant spread of hoax news in social media, especially in the political domain, poses a significant challenge that requires immediate attention. To address this issue, automatic hoax news detection using machine learning-based artificial intelligence has emerged as a promising approach. With the approaching presidential election in Indonesia in 2024, the need for effective detection methods becomes even more pressing.This research focuses on proposing an efficient deep learning model for detecting political hoax news on Indonesian social media. Word2vec feature representation and three deep learning models – LSTM, CNN, and Hybrid CNN-LSTM – are evaluated to determine the most effective approach. Experimental results reveal that the CNN-LSTM hybrid model outperforms the others, achieving an accuracy of 96% in detecting hoax news on Indonesian social media in the political domain. By leveraging state-of-the-art deep learning techniques, particularly the CNN-LSTM hybrid model, this study contributes to the advancement of hoax news detection in Indonesia's political landscape. The findings underscore the importance of utilizing sophisticated machine learning methods to combat the spread of misinformation, particularly during crucial political events such as elections.

Escaping the neutralization effect of modality features fusion in multimodal Fake News Detection

Fake news spreads at unprecedented speeds through online social media, raising many concerns and negative impacts on a variety of domains. To control this issue, Fake News Detection (FND) naturally becomes the chief task while multimodal FND has recently attracted more attention due to the fast-growing multimodal content in online social media. Commonly, the existing multimodal FND methods directly fuse the unimodal features into the multimodal features before detecting the veracity, however, such methods result in the neutralization effect problem caused by the veracity conflict between different modalities. To alleviate this problem, we propose a novel multimodal FND method, namely MultImodal fake NEws detectoR with Unimodal Veracity Signals (Miner-uvs). The basic idea of Miner-uvs is to incorporate an unimodal veracity classifier that can predict the veracity labels of each modality, and then use the unimodal veracity labels to weigh the unimodal features, so as to achieve more discriminative multimodal features. Because the unimodal veracity labels are unknown, we formulate the unimodal veracity classification as the Positive and Unlabeled (PU) learning problem; and solve it by using a variational PU learning method with a contrastive multimodal alignment regularization. To evaluate the effectiveness of Miner-uvs, we compare it with dozens of existing multimodal FND methods across benchmark datasets. Experimental results demonstrate that Miner-uvs consistently outperforms the existing competitors.

AI Assisted Deceptive Content Analysis System Using Bi-LSTM

Abstract: In an age of widespread misinformation, creating accurate machine learning algorithms is vital for ensuring the integrity of information sharing. This project aims to develop a machine learning model capable of distinguishing between fake and true news articles. [1]The model utilizes a Bidirectional Long Short-Term Memory (Bi-LSTM) neural network architecture. Exploratory data analysis techniques are employed to gain in-sights into the characteristics and distributions within the dataset. Data visualization techniques aid in understanding patterns and relationships within the dataset. Additionally, unigram analysis is conducted to extract meaningful features from the text data. The datasets are then prepared for model training, involving preprocessing steps such as tokenization and vectorization. Fi-nally, the Bi-LSTM model is constructed, leveraging its ability to capture long-range dependencies in sequential data. The model is trained on the prepared datasets, optimized using appropri-ate techniques, and evaluated using metrics such as accuracy, precision, recall, and F1-score. The Bi-LSTM architecture offers the advantage of capturing long-range dependencies in sequential data, thereby enhancing the model’s ability to discern nuanced patterns in news articles. The primary objective of this project is to develop a robust and accurate system for automatically detecting fake news, thus playing a pivotal role in enhancing the dissemination of reliable information in the digital age.

Data & Semantic Analysis Fake News Detection Based Domain-Specific Auto Encoder Learning

Abstract: Efforts to create technology for automatically detecting fake news are actively underway as the spread of misinformation on social media continues to escalate. However, most of these approaches primarily concentrate on the linguistic and structural aspects of fake news, such as identifying sources or authors, message length, and the frequency of negative language. In contrast, our study introduces a phony news detection model leveraging machine learning that incorporates user behaviors, news, and social capital-based social network dynamics. We used the XG Boost model to evaluate each feature's significance and identify the critical elements affecting the identification of fake news to capture the variables related to the spread of phony news entirely. We implemented SVM, RF, LR, CART, and NNET, well-known machine learning classification models, using these identified factors and assessed how well they detected bogus news. This work used a cross-validation process to prevent overfitting and to generalize the established models. Additionally, the predicted accuracy of the established models was compared. The RF model had the best forecast accuracy, about 94%, while the NNET had the worst performance rate, about 92.1%. As disinformation is created and disseminated with growing sophistication, the results of this study should improve the effectiveness of false news detection systems.

Fake News Detection and Analysis

Abstract: The internet is filled with inaccurate information, which makes it tough to distinguish between what is true and what is not. This review paper explores how scholars and IT professionals are countering false information by examining various strategies. We start by looking at different types of fake news and how they circulate. We then discuss a range of methods and tools that can be used to identify false information, including advanced computer programs and language analysis. These approaches can be categorized into three groups: analysing the language used, taking into account contextual data, and studying the distribution trends on the internet. We also discuss a method for teaching computers to detect patterns using machine learning false narratives by learning to differentiate between emotions and language subtleties. The paper also emphasizes the significance of fairness and impartiality in detection technologies. We conclude by discussing possible future research areas for academics and technology professionals it's critical to stay current with the quickly developing field of fake news.