%0 Conference Paper %1 dinan2019queens %A Dinan, Emily %A Fan, Angela %A Williams, Adina %A Urbanek, Jack %A Kiela, Douwe %A Weston, Jason %B EMNLP %D 2019 %K bias dialogue gender generation nlp %T Queens are Powerful too: Mitigating Gender Bias in Dialogue Generation. %U http://arxiv.org/abs/1911.03842 %X Models often easily learn biases present in the training data, and their predictions directly reflect this bias. We analyze gender bias in dialogue data, and examine how this bias is actually amplified in subsequent generative chit-chat dialogue models. We measure gender bias in six existing dialogue datasets, and focus on the most biased one, the multi-player text-based fantasy adventure dataset LIGHT, as a testbed for our bias mitigation techniques. The LIGHT dataset is highly imbalanced with respect to gender, containing predominantly male characters, likely because it is entirely collected by crowdworkers and reflects common biases that exist in fantasy or medieval settings. We consider three techniques to mitigate gender bias: counterfactual data augmentation, targeted data collection, and bias controlled training. We show that our proposed techniques mitigate gender bias in LIGHT by balancing the genderedness of generated dialogue utterances and are particularly effective in combination. We quantify performance using various evaluation methods---such as quantity of gendered words, a dialogue safety classifier, and human studies---all of which show that our models generate less gendered, but equally engaging chit-chat responses.

@inproceedings{dinan2019queens, abstract = {Models often easily learn biases present in the training data, and their predictions directly reflect this bias. We analyze gender bias in dialogue data, and examine how this bias is actually amplified in subsequent generative chit-chat dialogue models. We measure gender bias in six existing dialogue datasets, and focus on the most biased one, the multi-player text-based fantasy adventure dataset LIGHT, as a testbed for our bias mitigation techniques. The LIGHT dataset is highly imbalanced with respect to gender, containing predominantly male characters, likely because it is entirely collected by crowdworkers and reflects common biases that exist in fantasy or medieval settings. We consider three techniques to mitigate gender bias: counterfactual data augmentation, targeted data collection, and bias controlled training. We show that our proposed techniques mitigate gender bias in LIGHT by balancing the genderedness of generated dialogue utterances and are particularly effective in combination. We quantify performance using various evaluation methods---such as quantity of gendered words, a dialogue safety classifier, and human studies---all of which show that our models generate less gendered, but equally engaging chit-chat responses.}, added-at = {2021-01-25T14:08:56.000+0100}, author = {Dinan, Emily and Fan, Angela and Williams, Adina and Urbanek, Jack and Kiela, Douwe and Weston, Jason}, biburl = {https://www.bibsonomy.org/bibtex/23f3302b3d695497ff509cac30a49a62d/schwemmlein}, booktitle = {EMNLP}, description = {Queens are Powerful too: Mitigating Gender Bias in Dialogue Generation}, interhash = {bba1167d7e47317e328c685fe544f8ba}, intrahash = {3f3302b3d695497ff509cac30a49a62d}, keywords = {bias dialogue gender generation nlp}, note = {cite arxiv:1911.03842}, timestamp = {2021-01-25T14:08:56.000+0100}, title = {Queens are Powerful too: Mitigating Gender Bias in Dialogue Generation.}, url = {http://arxiv.org/abs/1911.03842}, year = 2019 }