Detecting Fake News Using Machine Learning

PurposeThe creation and dissemination of fake news can have severe consequences for a company’s brand. Researchers, policymakers and practitioners are eagerly searching for solutions to get us out of the “fake news crisis”. Here, one approach is to use automated tools, such as artificial intelligence (AI) algorithms, to support managers in identifying fake news. The study in this paper demonstrates how AI with its ability to analyze vast amounts of unstructured data, can help us tell apart fake and real news content. Using an AI application, this study examines if and how the emotional appeal, i.e., sentiment valence and strength of specific emotions, in fake news content differs from that in real news content. This is important to understand, as messages with a strong emotional appeal can influence how content is consumed, processed and shared by consumers.Design/methodology/approachThe study analyzes a data set of 150 real and fake news articles using an AI application, to test for differences in the emotional appeal in the titles and the text body between fake news and real news content.FindingsThe results suggest that titles are a strong differentiator on emotions between fake and real news and that fake news titles are substantially more negative than real news titles. In addition, the results reveal that the text body of fake news is substantially higher in displaying specific negative emotions, such as disgust and anger, and lower in displaying positive emotions, such as joy.Originality/valueThis is the first empirical study that examines the emotional appeal of fake and real news content with respect to the prevalence and strength of specific emotion dimensions, thus adding to the literature on fake news identification and marketing communications. In addition, this paper provides marketing communications professionals with a practical approach to identify fake news using AI.

Fake news occurs when a news article is intentionally released and subsequently shown to be false. This can negatively impact on an individual’s or indeed society’s comprehension of topics and shape public opinion on major events, e.g. elections. With the ubiquity of social media and fake news websites, people are increasingly exposed to fake news. In this article, we explore and compare the performance of fake news detection approaches using both machine and deep learning methods. Specifically we explore Logistic Regression, Support Vector Machines (SVM), Recurrent Neural Networks (RNN), Long Short Term Memory (LSTM) and Bidirectional Encoder Representations from Transformers (BERT) and compare the results with human crowds ability to distinguish real vs fake news. We utilise over 8.5m records from FakeNewsCorpus and 23k records from FakeNewsNet. The results shown that BERT achieved the best overall accuracy at 82.5%. Groups of individual assessors achieved an accuracy of 79%, whilst individuals varied significantly between 60-80% in their ability to distinguish real vs fake news.

The rise of fake news has become a significant global concern, undermining public trust and information integrity. This study explores the application of advanced machine learning algorithms for detecting fake news, leveraging a balanced dataset of real and fake news articles. Through rigorous preprocessing, including text cleaning and Term Frequency-Inverse Document Frequency (TF-IDF) vectorization, the study enhances data quality and model performance. Five machine learning models—Random Forest, Support Vector Machine (SVM), Neural Networks, Logistic Regression, and Naïve Bayes—are systematically evaluated using metrics such as accuracy, precision, recall, and F1-score. Results indicate that the Random Forest Classifier outperforms other models with an accuracy of 99.95% and balanced performance across metrics, demonstrating its robustness in distinguishing fake from real news. SVM and Neural Networks also achieve high accuracy, showcasing their capability in handling complex data. Logistic Regression and Naïve Bayes, while computationally efficient, exhibit relatively lower performance. The findings underscore the importance of ensemble methods and sophisticated preprocessing techniques in detecting fake news effectively. This research provides a methodological framework for scalable fake news detection, offering valuable insights for developing automated systems to combat misinformation and promote informed decision-making in the digital age.

Filtering, vetting, and verifying digital information is an area of core interest in information science. Online fake news is a specific type of digital misinformation that poses serious threats to democratic institutions, misguides the public, and can lead to radicalization and violence. Hence, fake news detection is an important problem for information science research. While there have been multiple attempts to identify fake news, most of such efforts have focused on a single modality (e.g., only text‐based or only visual features). However, news articles are increasingly framed as multimodal news stories, and hence, in this work, we propose a multimodal approach combining text and visual analysis of online news stories to automatically detect fake news. Drawing on key theories of information processing and presentation, we identify multiple text and visual features that are associated with fake or credible news articles. We then perform a predictive analysis to detect features most strongly associated with fake news. Next, we combine these features in predictive models using multiple machine‐learning techniques. The experimental results indicate that a multimodal approach outperforms single‐modality approaches, allowing for better fake news detection.

This paper argues that the harm of fake news lies partly in its inequality-reinforcing effect. It shows how fake news interacts with two background inequalities, namely the unequal distribution of opportunities to become competent knowers and the unequal social status of oppressed minorities. First, I argue that fake news reinforces epistemic inequality among its recipients. The epistemic harm of believing fake news does not uniformly affect all members of society. Instead, epistemic inequality makes some citizens more vulnerable to believing fake news. In turn, believing fake news amplifies epistemic inequality by making those who fall for fake news increasingly less competent as knowers. Second, I argue that fake news amplifies status inequality within society by downgrading the status of minorities who are negatively represented in fake news stories. A significant subset of fake news targets minorities who are not recognized as equal within society, such as immigrants, Muslims and Roma minorities. The harm of being negatively portrayed in anti-minority fake news stories falls asymmetrically on members of the targeted minorities, reinforcing their oppression. Focusing on anti-immigrant fake news as the most widespread and paradigmatic example of anti-minority fake news, I show how it amplifies cultural imperialism, marginalization and systemic violence against members of immigrant minorities. Anti-minority fake news is also specifically harmful to members of targeted minorities as political agents because it contributes to denying their equal status as members of the democratic political community. This subset of fake news stories is therefore particularly corrosive to the democratic system and its core values.

The availability and expansion of social media has made it difficult to distinguish between fake and real news. Information falsification has exponentially increased as a result of how simple it is to spread information through sharing. Social media dependability is also under jeopardy due to the extensive dissemination of false information. Therefore, it has become a research problem to automatically validate information, specifically source, content, and publisher, to identify it as true or false. Despite its limitations, machine learning (ML) has been crucial in the categorization of information. Previous studies suggested three-step methods for categorising false information on social media. In the first step of the process, the data set is subjected to a number of pre-processing processes in order to transform unstructured data sets into structured data sets. The unknowable properties of fake news and the features are extracted by the Lexicon Model in the second stage. In the third stage of this research project, a feature selection method by WOA (Whale Optimization Algorithm) for weight value to tune the classification part. Finally, a Hybrid Classification model that is hybrid with a fuzzy based Convolutional Neural Network and kernel based support vector machine is constructed in order to identify the data pertaining to bogus news. However using single classifier for fake news detection produces the insufficient accuracy. To overcome this issue in this work introduced an improved model for fake news classification. To turn unstructured data sets into structured data sets, a variety of pre-processing operations are used on the data set in the initial phase of the procedure. The unknowable properties of fake news and the features are extracted by the Lexicon Model in the second stage. In the third stage of this research project, a feature selection method by COA (Coati Optimization Algorithm) for weight value to tune the classification part. Finally, an ensemble of RNN (Recurrent Neural Networks), VGG-16 and ResNet50.A classification model was developed to recognise bogus news information. Evaluate each fake news analysis' performance in terms of accuracy, precision, recall, and F1 score. The suggested model, out of all the methodologies taken into consideration in this study, provides the highest outcomes, according to experimental findings

Nowadays, the dissemination of news information has become more rapid, liberal, and open to the public. People can find what they want to know more and more easily from a variety of sources, including traditional news outlets and new social media platforms. However, at a time when our lives are glutted with all kinds of news, we cannot help but doubt the veracity and legitimacy of these news sources; meanwhile, we also need to guard against the possible impact of various forms of fake news. To combat the spread of misinformation, more and more researchers have turned to natural language processing (NLP) approaches for effective fake news detection. However, in the face of increasingly serious fake news events, existing detection methods still need to be continuously improved. This study proposes a modified proof-of-concept model named NER-SA, which integrates natural language processing (NLP) and named entity recognition (NER) to conduct the in-domain and cross-domain analysis of fake news detection with the existing three datasets simultaneously. The named entities associated with any particular news event exist in a finite and available evidence pool. Therefore, entities must be mentioned and recognized in this entity bank in any authentic news articles. A piece of fake news inevitably includes only some entitlements in the entity bank. The false information is deliberately fabricated with fictitious, imaginary, and even unreasonable sentences and content. As a result, there must be differences in statements, writing logic, and style between legitimate news and fake news, meaning that it is possible to successfully detect fake news. We developed a mathematical model and used the simulated annealing algorithm to find the optimal legitimate area. Comparing the detection performance of the NER-SA model with current state-of-the-art models proposed in other studies, we found that the NER-SA model indeed has superior performance in detecting fake news. For in-domain analysis, the accuracy increased by an average of 8.94% on the LIAR dataset and 19.36% on the fake or real news dataset, while the F1-score increased by an average of 24.04% on the LIAR dataset and 19.36% on the fake or real news dataset. In cross-domain analysis, the accuracy and F1-score for the NER-SA model increased by an average of 28.51% and 24.54%, respectively, across six domains in the FakeNews AMT dataset. The findings and implications of this study are further discussed with regard to their significance for improving accuracy, understanding context, and addressing adversarial attacks. The development of stylometric detection based on NLP approaches using NER techniques can improve the effectiveness and applicability of fake news detection.

Fake news is a growing problem on social media and can have significant negative consequences for individuals and society owing to its accessibility, low cost, and quick distribution. It is difficult to automatically identify bogus news that defies the current content-based analysis techniques. One of the key reasons is that current NLP algorithms still lack common sense, which is frequently necessary for understanding how to read the news. Recent research has demonstrated that the propagation patterns of true and fake news differ on social media. With the potential for automatic fake news detection, this study investigates the use of Deep Learning (DL) models to detect fake news on social media. A Neural Network (NN) model was developed using Natural Language Toolkit (NLTK), TensorFlow, and Natural Language Processing (NLP) for textual analysis. The model was trained on a dataset of fake and real news articles, and was able to achieve high accuracy in identifying fake news. The results of this study suggest that DL models can be valuable tools for detecting fake news on social media.

Head is Where the Herd is: Fake News and Effect of Online Social Conformity on Islamophobia in Indians

This project investigates how Artificial Intelligence (AI), specifically supervised machine learning techniques, can be applied to detect and classify fake news with high accuracy. The motivation stems from the widespread dissemination of misinformation on social media, where false narratives often spread faster than verified content. To address this issue, two distinct classification models were implemented and evaluated: one combining TF-IDF vectorization with Random Forest classifier, and another using TF-IDF with Logistic Regression. The TF-IDF technique was used to convert raw textual data into meaningful numerical features, capturing word frequency and relevance within the corpus. Both models were trained and tested on a balanced dataset of true and fake news articles. Their performance was compared based on key metrics such as accuracy, precision, recall, and F1-score. The findings revealed that the Random Forest model consistently outperformed Logistic Regression, particularly when trained on uncleaned (raw) text. This suggests that Random Forest is more robust to linguistic noise, making it a more suitable choice for deployment in real-time fake news detection systems, especially on user-generated content platforms like social media.

News currently spreads rapidly through the internet. Because fake news stories are designed to attract readers, they tend to spread faster. For most readers, detecting fake news can be challenging and such readers usually end up believing that the fake news story is fact. Because fake news can be socially problematic, a model that automatically detects such fake news is required. In this paper, we focus on data-driven automatic fake news detection methods. We first apply the Bidirectional Encoder Representations from Transformers model (BERT) model to detect fake news by analyzing the relationship between the headline and the body text of news. To further improve performance, additional news data are gathered and used to pre-train this model. We determine that the deep-contextualizing nature of BERT is best suited for this task and improves the 0.14 F-score over older state-of-the-art models.

A fake news story can travel half way across the world as the truth puts on its socks. There are myriads of challenges facing journalists in spotting fake news hence its wide proliferation. Fake news has become a prominent subject of enquiry especially following its alleged influence of the 2016 general elections in US. Unfortunately, research on fake news has focused on social media, politics, elections, and economies. Few studies have focused on the challenges that TV journalists face in spotting fake news prompting this study. The specific research question was; what are the challenges facing television journalists in spotting fake news in Kenya? The study adapted a relativist-constructivist/interpretivist ontology and epistemology, qualitative approach and multiple case study methodology. Data was generated through in-depth interviews, direct observation and documents review. The study used purposive sampling to generate data from 16 journalists. Data was then analysed in themes and presented in narrative form. Key findings were that in spotting fake news, journalists faced challenges like; loss of viewers, lack of authoritative contacts, sources who gave fake news for personal, business, political, and economic benefits, ability of fake news to camouflage real news, speed of fake news, typologies of fake news, live reporting, inexperienced correspondents and interns, and social media. The study concludes that the challenges facing journalists in spotting fake news were majorly based on sources, technology, education, skills and training, and its typology. The study therefore recommends that editorial boards invest in experts to train journalists on styles, architecture, propagation and use of fake news, inoculation of journalists and audiences, raising fake news literacy levels, and use of technology based approaches like reverse search and fact checking sites.
 Key words: Fake news, journalists, spotting, challenges, television, Kenya

Individual factors such as cognitive capacities matter when one is requested to spot fake news. We suggest, however, that social influence—specifically as exercised by an authoritarian leader—might matter more if one is expected to agree with the fake news. We developed a single-item prototype measure of leadership styles and recruited participants from four Western democratic countries (Australia, Canada, United Kingdom, United States, N = 501) who identified their immediate boss as an autonomous, paternalistic, or authoritarian leader. Then they were asked to evaluate the accuracy of several fake news articles and their expectations to agree with their boss when asked about these articles. People with authoritarian bosses were less accurate in spotting fake news (Cohen’s d = 0.32) compared to employees with autonomous bosses. The bigger effect, however, was that they would agree with their boss about the fake news article when it was shared by their authoritarian boss compared to employees with autonomous (Cohen’s d = 1.30) or paternalistic bosses (Cohen’s d = 0.70). We argue that in addition to effects on the perceived accuracy of information, social influence, conformity, and obedience are crucial and unacknowledged factors of how misinformation may be maintained and propagated by authoritarian leaders.

In order to categorize healthy and stressed potato crops obtained from UAV-based multispectral photos, this study will use machine learning. It also aims to determine which spectral band offers the best separation for classification. These machine learning techniques are compared to classify healthy and stressed crops: Random Forest Classifier (RF), K-Nearest Neighbors Classifier (KNN), and Gradient Boosting Classifier (GBC), as well as Support Vector Classifier (SVC) and Linear Support Vector Classifier (LSVC). The Normalized Difference Vegetation Index (NDVI) and the Red Edge Normalized Difference Vegetation Index (NDRE), RED, GREEN, Near-Infrared (NIR), REDEDGE bands, are used by machine learning techniques to categorize data. The suggested approach consists of three main parts: (1) extraction of vegetation indicators; (2) Utilising thresholding technique to remove the background containing weeds and soil; and (3) classification and model training using machine learning classifier. As a consequence of its strong performance and good accuracy (0.777% accuracy and AUC score of 0.92), the paper’s findings demonstrate that RF is the best appropriate classifier for healthy and unhealthy plants. The NIR band is the most accurate in spotting unhealthy plants, with a 0.169 feature significance score. This band is connected to the two vegetative indices, NDVI and NDR, which had the best spectral characteristics in the classification model, with feature significance scores of 0.189 and 0.181, respectively.

The dissemination of fake news in an online social network platform has been a real concern nowadays. Through social media, news articles are posted by many sources like news channels, websites, or even newspaper websites. There is a need to be sure that the information posted is only from credible sources and these posts have to authenticate. The intensity of genuineness of the news posted online cannot be measured absolutely and is still challenging. A novel Score -based Multi-Source Fake News Detection framework is proposed in this work to automate the detection of fake news from multiple news sources. This framework extracts the text-based features from genuine and fake news articles using Term Frequency - Inverted Document Frequency. Then the credibility score of the source s is calculated based on the site_url features and Top Level Domain. By assimilating the text-based features with the credibility score of multi-source, the credibility of the news is estimated. The proposed framework is applied to the Machine Learning (ML) classifiers to examine their performance in the detection of fake news. The experimental results determine the efficacy of the proposed framework with the Gradient Boosting algorithm of about 99.5% to the utmost level.