Early Detection of Rumor Veracity in Social Media
Bibliographic record
Abstract
Rumor spread has become a significant issue in online social networks (OSNs). To mitigate and limit the spread of rumors and its detrimental effects, analyzing, detecting and better understanding rumor dynamics is required. One of the critical steps of studying rumor spread is to identify the level of the rumor truthfulness in its early stage. Understanding and identifying the level of rumor truthfulness helps prevent its viral spread and minimizes the damage a rumor may cause. In this research, we aim to debunk rumors by analyzing, visualizing, and classifying the level of rumor truthfulness from a large number of users that actively engage in rumor spread. First, we create a dataset of rumors that belong to one of five categories: "False", "Mostly False", "True", "Mostly True", and "Half True". This dataset provides intrinsic characteristics of a rumor: topics, user's sentiment, network structural and content features. Second, we analyze and visualize the characteristics of each rumor category to better understand its features. Third, using theories from social science and psychology, we build a feature set to classify those rumors and identify their truthfulness. The evaluation results on our new dataset show that the approach could effectively detect the truth of rumors as early as seven days. The proposed approach could be used as a valuable tool for existing fact-checking websites, such as Snopes.com or Politifact.com, to detect the veracity of rumors in its early stage automatically and educate OSN users to have a well-informed decision-making process.
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 | 0.006 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.001 | 0.004 |
| Open science | 0.007 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| 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".