Fairness Through Controlled (Un)Awareness in Node Embeddings
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914891525259264 |
|---|---|
| author | Vetter, Dennis Forth, Jasper Roig, Gemma Dell, Holger |
| author_facet | Vetter, Dennis Forth, Jasper Roig, Gemma Dell, Holger |
| contents | Graph representation learning is central for the application of machine learning (ML) models to complex graphs, such as social networks. Ensuring `fair' representations is essential, due to the societal implications and the use of sensitive personal data. In this paper, we demonstrate how the parametrization of the \emph{CrossWalk} algorithm influences the ability to infer a sensitive attributes from node embeddings. By fine-tuning hyperparameters, we show that it is possible to either significantly enhance or obscure the detectability of these attributes. This functionality offers a valuable tool for improving the fairness of ML systems utilizing graph embeddings, making them adaptable to different fairness paradigms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_20024 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Fairness Through Controlled (Un)Awareness in Node Embeddings Vetter, Dennis Forth, Jasper Roig, Gemma Dell, Holger Social and Information Networks Computers and Society Graph representation learning is central for the application of machine learning (ML) models to complex graphs, such as social networks. Ensuring `fair' representations is essential, due to the societal implications and the use of sensitive personal data. In this paper, we demonstrate how the parametrization of the \emph{CrossWalk} algorithm influences the ability to infer a sensitive attributes from node embeddings. By fine-tuning hyperparameters, we show that it is possible to either significantly enhance or obscure the detectability of these attributes. This functionality offers a valuable tool for improving the fairness of ML systems utilizing graph embeddings, making them adaptable to different fairness paradigms. |
| title | Fairness Through Controlled (Un)Awareness in Node Embeddings |
| topic | Social and Information Networks Computers and Society |
| url | https://arxiv.org/abs/2407.20024 |