Arabic Fake News Dataset Development: Humans and AI-Generated Contributions
Bibliographic record
Abstract
The extensive use of social media platforms has promoted the rapid spread of fake news on the internet, such as fake reviews, rumors, and propaganda. Although these terminologies have different objectives, they share the aim of causing harm in the form of fake news. This study presents an Arabic fake news detection framework to overcome the widespread fake news phenomenon. The proposed framework introduces the first Arabic fake news dataset compiled by passing through strict guidelines to produce fake articles composed by humans and the generative pre-trained transformer (GPT). First, we performed human-based experiments to evaluate the ability of humans to distinguish real news articles from fake news articles. Our findings reveal that humans could roughly identify half of the fake articles from humans or GPT, raising concerns about their ability to detect fake news. This highlights the growing concern surrounding fake news, especially because GPT demonstrates the ability to generate fake news that closely resembles human-created content, further amplifying the issue. To address this issue, we performed the same task using Deep Learning (DL) and transformer-based methods with different word embeddings. Across all the employed models, the study revealed that the innovative transformer-based model, ARBERT, outperformed the DL models, reaching an accuracy of 78% in classifying real and fake news generated by humans and GPT. The findings suggest effective techniques for addressing and resolving this issue.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".