The unprecedented scale of disinformation on the Internet for more than a decade represents a serious challenge for democratic societies. When this process is focused on a well-established subject such as climate change, it can subvert measures and policies that various governmental bodies have taken to mitigate the phenomenon. It is therefore essential to effectively identify and counteract fake news on climate change. To do this, our main contribution represents a novel dataset with more than 2300 articles written in French, gathered using web scraping from all types of media dealing with climate change. Manual labeling was performed by two annotators with three classes: “fake”, “biased”, and “true”. Machine Learning models ranging from bag-of-words representations used by an SVM to Transformer-based architectures built on top of CamemBERT were built to automatically classify the articles. Our results, with an F1-score of 84.75% using the BERT-based model at the article level coupled with hand-crafted features specifically tailored for this task, represent a strong baseline. At the same time, we highlight perceptual properties as text sequences (i.e., fake, biased, and irrelevant text fragments) at the sentence level, with a macro F1 of 45.01% and a micro F1 of 78.11%. Based on these results, our proposed method facilitates the identification of fake news, and thus contributes to better education of the public.

Fake news is one of the most debated social phenomena of recent years. It has been the subject not only of several attempts at defining it, but also of numerous comparative analyses of prevalent definitions. Nonetheless, the present article fosters the ambition of offering a new definition. The innovation of my definition stems from the fact that it departs from the dominant “hybrid view” on fake news, which considers the defining traits of the phenomenon to be its truth value (i.e., its falseness) and the intention of its author (i.e., to mislead its public). Opposing this view, the present article argues that the producer’s intent is irrelevant in regard to classifying news as fake news. On the contrary, the defining trait of fake news is, alongside the falsehood of its content, the discourse’s perlocutionary force, which invariably entails a call to action addressed to the text’s recipient.

This article sets out to offer an overview and a review of the latest linguistic research into fake news. To this end, the authors put forward a critical discussion of the paradigms and instruments deployed over the past decade to analyze and identify this textual (micro)genre, from natural language processing techniques to critical discourse analysis. The conclusion of our study is that a proper understanding of the fake news phenomenon can only be achieved by bringing together qualitative and quantitative methods.

Predefined thematic taxonomies can obscure how misinformation narratives emerge, overlap, and change over time. One way to address this limitation is by developing bottom-up data-driven approaches that identify narratives directly from data. We present a pipeline that extracts entities and relations from news articles using GraphRAG, identifies communities using the Leiden algorithm, and uses this information to explore and predict misinformation. We evaluate the pipeline on a set of Romanian COVID-19 pandemic news articles with expert-defined narrative annotations and adapt the extraction prompts to Romanian. Exploratory results show that in this corpus, the graph for fake news is more fragmented and less modular than the graph for real news; it also contains higher proportions of events, disinformation theory, and vaccine entities, whereas the real-news graph contains higher proportions of locations, hospitals, and public health measures. Communities in the combined graph align weakly with the predefined super-narrative labels, indicating that the two representations group articles differently. Furthermore, expert review of similar articles from communities identified within the combined real–fake graph suggests that an approach based on entity and community may offer a promising avenue for identifying subtle misinformation strategies. Regarding the prediction task, a three-component ensemble that uses graph information achieves 85.86% macro-F1 (95% CI [78.50, 91.27]) and 86.51% accuracy (95% CI [78.57, 91.27]) on classifying articles as fake or true. These point estimates suggest incremental gains over the two-component ensemble without the graph similarity signal, though these gains are not statistically confirmed at this sample size. Together, these exploratory results illustrate the use of GraphRAG as a tool in pipelines that support the exploration and prediction of misinformation.

This article demonstrates that some of the textemes of Romanian playwright I.L. Caragiale, which later became famous throughout the entire Romanian ethnolinguistic community, are in fact borrowed from older Romanian texts, which appeared in the mid-19th century. My analysis focuses on three textemes from Caragiale’s comedies (“the family is the small homeland, as the homeland is the large family”, “a stormy night” and “have a little patience”), identifying their origins in the writings of C. D. Aricescu, Ioan M. Bujoreanu and I. Heliade-Rădulescu. My study concludes that Caragiale’s plays have a deeply intertextual structure.