Contextual Brushing is a methodological contribution in computational text analysis that selectively highlights or downweights contextual elements within a dataset to refine the interpretability or performance of a downstream model.