|
--- |
|
language: "en" |
|
thumbnail: "https://pbs.twimg.com/profile_images/1092721745994440704/d6R-AHzj_400x400.jpg" |
|
tags: |
|
- propaganda |
|
- bert |
|
license: "MIT" |
|
datasets: |
|
- |
|
metrics: |
|
- |
|
--- |
|
|
|
Propaganda Techniques Analysis BERT |
|
---- |
|
|
|
This model is a BERT based model to make predictions of propaganda techniques in |
|
news articles in English. It was introduced in |
|
[this paper](https://propaganda.qcri.org/papers/EMNLP_2019__Fine_Grained_Propaganda_Detection.pdf). |
|
|
|
|
|
## Model description |
|
|
|
Please find propaganda definition here: |
|
https://propaganda.qcri.org/annotations/definitions.html |
|
|
|
|
|
## How to use |
|
|
|
```python |
|
>>> from .model import BertForTokenAndSequenceJointClassification |
|
>>> model = BertForTokenAndSequenceJointClassification.from_pretrained( |
|
>>> "QCRI/PropagandaTechniquesAnalysis-en-BERT", |
|
>>> revision="v0.1.0", |
|
>>> ) |
|
``` |
|
|