Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# MMM-Fact: A Multimodal, Multi-Domain Fact-Checking Dataset
|
| 2 |
+
|
| 3 |
+
## Overview
|
| 4 |
+
MMM-Fact is a real-world, multimodal benchmark for fact checking spanning 1995–2025. Each claim is paired with the full fact-check article and cross-modal evidence (text, images, videos, tables) to evaluate multi-step, cross-modal retrieval and reasoning.
|
| 5 |
+
|
| 6 |
+
## Scale & Coverage
|
| 7 |
+
| Item | Details |
|
| 8 |
+
|---|---|
|
| 9 |
+
| Instances | **125,449** fact-checked claims |
|
| 10 |
+
| Time range | **1995–2025** |
|
| 11 |
+
| Sources | **4** fact-checking sites + **1** news outlet *(each sample includes the complete fact-check article and associated evidence)* |
|
| 12 |
+
|
| 13 |
+
## Annotations & Tasks
|
| 14 |
+
- **Veracity labels:** True / False / Not Enough Information
|
| 15 |
+
- **Supported tasks:**
|
| 16 |
+
- Veracity prediction
|
| 17 |
+
- Explainable fact-checking
|
| 18 |
+
- Complex evidence aggregation
|
| 19 |
+
- Longitudinal analysis
|
| 20 |
+
|
| 21 |
+
## Retrieval-Difficulty Tiers
|
| 22 |
+
To reflect verification effort, each sample is tagged with a retrieval difficulty based on the number of retrievable evidence sources:
|
| 23 |
+
|
| 24 |
+
| Tier | Evidence Sources |
|
| 25 |
+
|---|---|
|
| 26 |
+
| **Basic** | 1–5 sources |
|
| 27 |
+
| **Intermediate** | 6–10 sources |
|
| 28 |
+
| **Advanced** | >10 sources |
|
| 29 |
+
|
| 30 |
+
## Citation
|
| 31 |
+
If you find this dataset useful, please cite: [https://arxiv.org/abs/2510.25120](https://arxiv.org/abs/2510.25120)
|