@unpublished{Lertvittayakumjorn:2020, author = {Lertvittayakumjorn, P and Specia, L and Toni, F}, title = {FIND: human-in-the-loop debugging deep text classifiers}, url = {http://arxiv.org/abs/2010.04987v1}, year = {2020} }
TY - UNPB AB - Since obtaining a perfect training dataset (i.e., a dataset which isconsiderably large, unbiased, and well-representative of unseen cases) ishardly possible, many real-world text classifiers are trained on the available,yet imperfect, datasets. These classifiers are thus likely to have undesirableproperties. For instance, they may have biases against some sub-populations ormay not work effectively in the wild due to overfitting. In this paper, wepropose FIND -- a framework which enables humans to debug deep learning textclassifiers by disabling irrelevant hidden features. Experiments show that byusing FIND, humans can improve CNN text classifiers which were trained underdifferent types of imperfect datasets (including datasets with biases anddatasets with dissimilar train-test distributions). AU - Lertvittayakumjorn,P AU - Specia,L AU - Toni,F PY - 2020/// TI - FIND: human-in-the-loop debugging deep text classifiers UR - http://arxiv.org/abs/2010.04987v1 UR - http://hdl.handle.net/10044/1/83501 ER -