Data Augmentation
Explore data augmentation strategies that improve fairness in natural language processing. Learn to generate diverse training examples through swapping gendered words and paraphrasing, helping to create more unbiased models by presenting existing data in varied ways.
We'll cover the following...
Augmentation
Modifying word vectors to enhance fairness is a good strategy. However, its application is restricted to static vectors. What if we leverage contextual embeddings like BERT? Or refrain from using vectors altogether? We dig into a method that is universally applicable.
Data augmentation involves generating additional training examples. These examples aim to diversify the training data, assisting the model in discerning more robust relationships. We might wonder: by deriving augmented data from the training set, do we genuinely enrich the information available to the model? The answer is yes. The perspective here is not to create new information out of nowhere but to present the existing data in a diverse manner. ...