Unsupervised detection of contextualized embedding bias with application to ideology
We propose a fully unsupervised method to detect bias in contextualized embeddings. The method leverages the assortative information latently encoded by social networks and combines orthogonality regularization, structured sparsity learning, and graph neural networks to find the embedding subspace c...
Príomhchruthaitheoirí: | Hofmann, V, Pierrehumbert, J, Schütze, H |
---|---|
Formáid: | Conference item |
Teanga: | English |
Foilsithe / Cruthaithe: |
Journal of Machine Learning Research
2022
|
Míreanna comhchosúla
Míreanna comhchosúla
-
Dynamic contextualized word embeddings
de réir: Hofmann, V, et al.
Foilsithe / Cruthaithe: (2021) -
Leveraging Bias in Pre-trained Word Embeddings for Unsupervised Microaggression Detection
de réir: Tolúlọpẹ́ Ògúnrẹ̀mí, et al.
Foilsithe / Cruthaithe: (2022-12-01) -
Unsupervised Bitext Mining and Translation via Self-Trained Contextual Embeddings
de réir: Phillip Keung, et al.
Foilsithe / Cruthaithe: (2021-03-01) -
Forecasting COVID-19 caseloads using unsupervised embedding clusters of social media posts
de réir: Drinkall, F, et al.
Foilsithe / Cruthaithe: (2022) -
Modeling ideological salience and framing in polarized online groups with graph neural networks and structured sparsity
de réir: Hofmann, V, et al.
Foilsithe / Cruthaithe: (2022)