Research into fake news detection has a long history, although it gained significant attention following the 2016 US election. During this time, the widespread use of social media and the resulting increase in interpersonal communication led to the extensive spread of ambiguous and potentially misleading news. Traditional approaches, relying solely on pre- LLM (large language model) techniques and addressing the issue as a simple classification problem, have shown to be insufficient for improving detection accuracy. In this context, large language models have become crucial, as their advanced architectures overcome the limitations of pre-LLM methods, which often fail to capture the subtleties of fake news. This literature review aims to shed light on the field of fake news detection by providing a brief historical overview, defining fake news, reviewing detection methods used before the advent of large language models, and discussing the strengths and weaknesses of these models in an increasingly complex landscape. Furthermore, it will emphasize the importance of using multimodal datasets in the effort to detect fake news.

This study compares Transformer-based models and Graph Neural Networks (GNNs) for fake news detection across three datasets: FakeNewsNet, ISOT, and WELFake. Transformer models (BERT, RoBERTa, GPT-2) demonstrated superior performance, achieving mean accuracies above 85% on FakeNewsNet and exceeding 98% on ISOT and WELFake. Specifically, RoBERTa achieved 86.16% accuracy on FakeNewsNet and 99.99% on ISOT, while GPT-2 reached 99.72% on WELFake. In contrast, GNNs (GCN, GraphSAGE, GIN, GAT) exhibited lower performance. GCN achieved 71% accuracy on FakeNewsNet but dropped to 53.30% on ISOT and 50.28% on WELFake, with F1 scores reflecting similar trends. Other GNNs, like GraphSAGE, showed even lower results, particularly on ISOT and WELFake, where performance hovered around 50%. Our findings indicate that while Transformers provide exceptional accuracy and reliability, GNNs offer potential efficiency benefits for resource-constrained scenarios despite their lower predictive performance. This study informs model selection for fake news detection tasks and encourages the exploration of hybrid approaches to balance accuracy and computational efficiency.

This brief survey acts as a fundamental resource for researchers beginning their exploration into fake news detection. It emphasizes the importance of dataset quality and diversity in enhancing the effectiveness of detection models, detailing key features, labeling systems, and prevalent biases. It also presents the challenges and limitations. By addressing ethical considerations (such as privacy and consent, societal impacts, transparency, and accountability) and best practices (annotation methodologies, real-world dynamics, reliability, and validity), we offer a thorough overview of current datasets. Additionally, our contribution includes a GitHub repository that aggregates publicly available datasets into a single, easily accessible portal, thereby supporting further research and development in the fight against fake news.

Clickbait headlines degrade the quality of online information and undermine user trust. We present a hybrid approach to clickbait detection that combines transformer-based text embeddings with linguistically motivated informativeness features. Using natural language processing techniques, we evaluate classical vectorizers, word embedding baselines, and large language model embeddings paired with tree-based classifiers. Our best-performing model, XGBoost over embeddings augmented with 15 explicit features, achieves an F1-score of 91\%, outperforming TF-IDF, Word2Vec, GloVe, LLM prompt based classification, and feature-only baselines. The proposed feature set enhances interpretability by highlighting salient linguistic cues such as second-person pronouns, superlatives, numerals, and attention-oriented punctuation, enabling transparent and well-calibrated clickbait predictions. We release code and trained models to support reproducible research.