Casual Conversations v2: Designing a large consent-driven dataset to measure algorithmic bias and robustness.
Caner HazirbasYejin BangTiezheng YuParisa AssarBilal PorgaliVítor AlbieroStefan HermanekJacqueline PanEmily McReynoldsMiranda BogenPascale FungCristian Canton-FerrerPublished in: CoRR (2022)
Keyphrases
- data driven
- similarity measure
- distance measure
- benchmark datasets
- neural network
- data sets
- information systems
- object recognition
- high robustness
- highly correlated
- information content
- correlation coefficient
- text classification
- user interface
- feature selection
- learning algorithm
- genetic algorithm
- information retrieval