Update README.md
Browse files
README.md
CHANGED
|
@@ -28,6 +28,8 @@ Dataset for the paper "[Beyond Truth Discovery: A Two-Stage Framework to Assess
|
|
| 28 |
| `harm_mv` | Harmfulness label from majority voting by three independent annotators: `1` = low, `2` = high. |
|
| 29 |
| `severity` | Combination of `believe_mv` and `harm_mv`, written in that order: `11`, `12`, `21`, or `22`. |
|
| 30 |
|
|
|
|
|
|
|
| 31 |
Believability measures how likely a claim is to be perceived as credible.
|
| 32 |
|
| 33 |
Harmfulness measures its potential consequences *if believed*.
|
|
@@ -36,8 +38,8 @@ Severity is derived from these two labels; it is not a separate annotation.
|
|
| 36 |
|
| 37 |
## Citation
|
| 38 |
|
| 39 |
-
|
| 40 |
-
author = {
|
| 41 |
title = {Beyond Truth Discovery: A Two-Stage Framework to Assess the Severity of False Claim during Disasters},
|
| 42 |
year = {2026},
|
| 43 |
isbn = {9798400728945},
|
|
|
|
| 28 |
| `harm_mv` | Harmfulness label from majority voting by three independent annotators: `1` = low, `2` = high. |
|
| 29 |
| `severity` | Combination of `believe_mv` and `harm_mv`, written in that order: `11`, `12`, `21`, or `22`. |
|
| 30 |
|
| 31 |
+
Only 300 randomly sampled claims have been annotated by human annotators for both hurricane and wildfire.
|
| 32 |
+
|
| 33 |
Believability measures how likely a claim is to be perceived as credible.
|
| 34 |
|
| 35 |
Harmfulness measures its potential consequences *if believed*.
|
|
|
|
| 38 |
|
| 39 |
## Citation
|
| 40 |
|
| 41 |
+
@inproceedings{FSD,
|
| 42 |
+
author = {Yao, Ruichen and Dasari, Tejna and Baispay, Gulshat and Zaurbek, Aizhan and Liu, Yifan and Liu, Yaokun and Li, Zelin and Wang, Dong},
|
| 43 |
title = {Beyond Truth Discovery: A Two-Stage Framework to Assess the Severity of False Claim during Disasters},
|
| 44 |
year = {2026},
|
| 45 |
isbn = {9798400728945},
|