Методы устранения предвзятости
Related Papers
Word Embeddings (Also) Encode Human Personality Stereotypes
AbstractWord representations trained on text reproduce human implicit bias related to gender, race and age. Methods have been developed to remove such bias. Here, we present results that show that …...
Does Gender Matter? Towards Fairness in Dialogue Systems
AbstractRecently there are increasing concerns about the fairness of Artificial Intelligence (AI) in real-world applications such as computer vision and recommendations. For example, recognition algorithms in computer vision are unfair …...
End-to-End Bias Mitigation by Modelling Biases in Corpora
AbstractSeveral recent studies have shown that strong natural language understanding (NLU) models are prone to relying on unwanted dataset biases without learning the underlying task, resulting in models that fail …...
Improving QA Generalization by Concurrent Modeling of Multiple Biases
AbstractExisting NLP datasets contain various biases that models can easily exploit to achieve high performances on the corresponding evaluation sets. However, focusing on dataset-specific biases limits their ability to learn …...
Look at the First Sentence: Position Bias in Question Answering
AbstractMany extractive question answering models are trained to predict start and end positions of answers. The choice of predicting answers as positions is mainly due to its simplicity and effectiveness. …...
Masking Actor Information Leads to Fairer Political Claims Detection
AbstractA central concern in Computational Social Sciences (CSS) is fairness: where the role of NLP is to scale up text analysis to large corpora, the quality of automatic analyses should …...
Mitigating Gender Bias for Neural Dialogue Generation with Adversarial Learning
AbstractDialogue systems play an increasingly important role in various aspects of our daily life. It is evident from recent research that dialogue systems trained on human conversation data are biased. …...
Towards Debiasing NLU Models from Unknown Biases
AbstractNLU models often exploit biases to achieve high dataset-specific performance without properly learning the intended task. Recently proposed debiasing methods are shown to be effective in mitigating this tendency. However, …...
Challenges in Automated Debiasing for Toxic Language Detection
AbstractBiased associations have been a challenge in the development of classifiers for detecting toxic language, hindering both fairness and accuracy. As potential solutions, we investigate recently introduced debiasing methods for …...
RedditBias: A Real-World Resource for Bias Evaluation and Debiasing of Conversational Language Models
AbstractText representation models are prone to exhibit a range of societal biases, reflecting the non-controlled and biased nature of the underlying pretraining data, which consequently leads to severe ethical issues …...
Are All Spurious Features in Natural Language Alike? An Analysis through a Causal Lens
AbstractThe term ‘spurious correlations’ has been used in NLP to informally denote any undesirable feature-label correlations. However, a correlation can be undesirable because (i) the feature is irrelevant to the …...
Debiasing Isn’t Enough! – on the Effectiveness of Debiasing MLMs and Their Social Biases in Downstream Tasks
AbstractWe study the relationship between task-agnostic intrinsic and task-specific extrinsic social bias evaluation measures for MLMs, and find that there exists only a weak correlation between these two types of …...