|
--- |
|
license: mit |
|
datasets: |
|
- koliskos/fake_news |
|
language: |
|
- en |
|
--- |
|
# Model Card for Model ID |
|
|
|
Model is used to detect whether a news story is fake or legitimate. |
|
|
|
|
|
|
|
- **Developed by:** koliskos |
|
- **Model type:** Text Classification |
|
- **Language(s) (NLP):** English |
|
- **License:** mit |
|
- **Finetuned from model:** DistilBERT |
|
|
|
|
|
- **Repository:** koliskos/fine_tuned_fake_news_classifier |
|
|
|
## Uses |
|
|
|
This model is meant to classify news articles as real or fake. |
|
|
|
## Bias, Risks, and Limitations |
|
|
|
This model could potentially assume "fake" to be the default |
|
prediction for news stories that contain names that are seen |
|
heavily within fake news articles, ex: a news story about someone |
|
named Hillary may be labeled fake even if it is real because the |
|
name Hillary is heavily grounded within the context of Hillary Clinton. |
|
|
|
|
|
## Model Card Contact |
|
|
|
spkolisko "at" wellesley.edu |