Jul 2026· Anais do XV Brazilian Workshop on Social Network Analysis and Mining (BraSNAM 2026)· pp. 70-82· 0 citations· 22 references
TL;DR
This paper proposes modeling data through ego networks, extracting features from the underlying network, matching with textual features, and using traditional machine learning algorithms to detect and identify fake news on social networks.
Abstract
This paper addresses the detection of fake news on social networks by combining complex networks and artificial intelligence techniques. Recent works have shown progress in solving the problem of detecting fake news using deep learning, which, in general, are penalized by the lack of interpretability and require large amounts of labeled data. In addition, to represent instances, solutions in the literature generally use textual characteristics, social relationships, and information related to engagement on social media. However, there are still gaps to be explored regarding the most relevant features of a fake post taking as a premise the interpretability of the solution. We propose modeling data through ego networks, extracting features from the underlying network, matching with textual features, and using traditional machine learning algorithms to detect and identify fake news on social networks. The experiments, carried out on Twitter data using the popular fake news dataset – FakeNewsNet, show the potential of the proposed approach from the perspectives of interpretability, precision and recall.
Social media has become usefull and most important platform for individuals to access news due to its speed and cost-effectiveness of disseminating information on a particular channel. However, these platforms also make it a breeding ground for the spread of fake news, which can impact society and individuals. With the advent of modern technology that comes with the fourth industrial revolution, promoting openness and increased engagement, social media has evolved to serve multiple purposes. It has become an integral part of our lives, beyond what was originally intended for just a small group of people. Consequently, identifying such fake news has become a crucial task for researchers and remains a major concern. This paper aims to examine the methods of publishing and distributing fake news, and outlines the approach of classifying, organizing, and developing algorithms. Our proposed solution is called the "Fake News Detection using Hybrid-kMeans (FNDHKM)" algorithm that utilizes Natural Language Processing (NLP) and machine learning (ML) techniques to detect fake news on social media platforms, specifically Twitter. The experiments were performed on three different datasets obtained from Twitter and involved dimensionality reduction on the extracted data. The highest accuracy achieved was 85% and 90% for precision and accuracy, respectively. The proposed FNDHKM approach showed significantly high accuracy with low overhead.
Shubha Mishra· International journal of com...· 0 citations
An Explainable Artificial Intelligence (XAI) framework for fake news detection that unites the complementary explainability methods: SHapley Additive exPlanations and Local Interpretable Model-agnostic Explanations with the feature extraction technique, Term Frequency-Inverse Document Frequency (TF-IDF) and the Linear Support Vector Machine (Linear SVM) classifier.
A multi-model learning framework that combines the complementary strengths of classical machine learning classifiers, deep sequential neural networks, and transformer-based contextual language models to detect fake news on social media is proposed.
Priya Verma· International Journal of Res...· 0 citations
Due to the fast proliferation of online news media and social networks, there is a significant increase in the dissemination of misleading data and fake news on digital platforms. Fake news detection is difficult due to the incapacity of deep learning models or conventional machine learning in comprehending semantic and higher-level interactions between news text, news author, publisher, and additional metadata. This results in decreased detection efficiency and enables the dissemination of misinformation that impacts societal behavior and public opinion. To address this problem, a Four-Level Hierarchical Attention Network (4HAN) that incorporates word-, sentence-, and headline-level attention, along with Hypergraph Convolution and Hypergraph Attention, is proposed using the LIAR dataset. By combining semantic feature extraction and relational dependency modeling, the 4HAN framework achieves better results. Results showed a detection accuracy rate of 96.00%, which beats multiple existing methodologies in fake news detection. These findings demonstrate that combining hierarchical attention with hypergraph learning provides more effective semantic and relational representation, leading to improved fake news detection performance and greater robustness in misinformation analysis.
Alpana A. Borse, Gajanan K. Kharate, N. Wasatkar· Journal of Intelligent Decis...· 0 citations
The majority of social media research relies on datasets gathered from social media websites such as Reddit and Twitter. However, their intrinsic high noisy content reduces the performance of data-driven models. This restriction of noisy data has made data preparation techniques necessary. Current systems usually need to pay more attention to data quality and their performance heavily depends on noisy social media datasets. In this work, we concentrate on creating high-quality social media data for rumor detection tasks on the widely popular PHEME-9 dataset. Eliminating noisy responses is essential for misinformation tasks since it guarantees that the model has been trained on precise and pertinent data, improving its capacity to identify and validate rumors successfully. Large language models (LLMs) are used in this work to filter out irrelevant comments prior to the application of machine and deep learning techniques. This bifurcated approach helps to improve model accuracy and lower computing burden. We assume that this will further help in accurate rumor identification and can support environmental sustainability using less computational resources. Our proposed methodology shows an average improvement in the trained filtered models’ performance in terms of accuracy and F1-scores on six events in the PHEME-9 dataset. Further, to validate the effectiveness of the trained model, we performed interpretability and error analysis.
Shakshi Sharma, Anjali Goyal, Naman Ahuja et al.· International Journal of Com...· 0 citations
Social media platforms are one of the most common ways people interact, exchange information and build online communities. Their rapid growth has been accompanied by a dramatic increase in the number of fraudulent user accounts. These fake profiles are frequently created for purposes such as identity theft, online fraud, phishing, spamming, spreading misinformation and influencing public opinion. Such activities threaten user confidence and raise security issues for users and social media companies. Hence the development of an automatic approach to detect fraudulent accounts has become increasingly important. In this study, we propose a machine learning based approach to detect fake profiles in social media. The proposed solution checks a number of profile attributes automatically to decide whether an account is real or fake, instead of relying on human verification. The key features such as number of followers, total number of posts, profile description, age of the account, the availability of the profile photo and other variables related to the account are used as input to the prediction process.
Sinchana B, Sharanya S Patel, Poornima H.A· International Journal of Res...· 0 citations
