Equal Protection Under the Algorithm: A Legal-Inspired Framework for Identifying Discrimination in Machine Learning. Soundarajan, S. & Clausen, D. L 2018. abstract bibtex Within the field of ethical machine learning, an area of special concern is the possibility of machine learning algorithms discriminating against groups of people in unethical ways, such as targeting advertisements based on race. In this paper, we propose a framework based on long-standing U.S. legal principles to determine whether the targeting of a group should be viewed with suspicion. Unlike existing work, we are focused on the case when the group is not correlated with known ‘protected features’, or such data is unavailable.
@article{soundarajan_equal_2018,
title = {Equal {Protection} {Under} the {Algorithm}: {A} {Legal}-{Inspired} {Framework} for {Identifying} {Discrimination} in {Machine} {Learning}},
abstract = {Within the field of ethical machine learning, an area of special concern is the possibility of machine learning algorithms discriminating against groups of people in unethical ways, such as targeting advertisements based on race. In this paper, we propose a framework based on long-standing U.S. legal principles to determine whether the targeting of a group should be viewed with suspicion. Unlike existing work, we are focused on the case when the group is not correlated with known ‘protected features’, or such data is unavailable.},
language = {en},
author = {Soundarajan, Sucheta and Clausen, Daniel L},
year = {2018},
}
Downloads: 0
{"_id":"BF9Yc5xh4sxNCJKtX","bibbaseid":"soundarajan-clausen-equalprotectionunderthealgorithmalegalinspiredframeworkforidentifyingdiscriminationinmachinelearning-2018","author_short":["Soundarajan, S.","Clausen, D. L"],"bibdata":{"bibtype":"article","type":"article","title":"Equal Protection Under the Algorithm: A Legal-Inspired Framework for Identifying Discrimination in Machine Learning","abstract":"Within the field of ethical machine learning, an area of special concern is the possibility of machine learning algorithms discriminating against groups of people in unethical ways, such as targeting advertisements based on race. In this paper, we propose a framework based on long-standing U.S. legal principles to determine whether the targeting of a group should be viewed with suspicion. Unlike existing work, we are focused on the case when the group is not correlated with known ‘protected features’, or such data is unavailable.","language":"en","author":[{"propositions":[],"lastnames":["Soundarajan"],"firstnames":["Sucheta"],"suffixes":[]},{"propositions":[],"lastnames":["Clausen"],"firstnames":["Daniel","L"],"suffixes":[]}],"year":"2018","bibtex":"@article{soundarajan_equal_2018,\n\ttitle = {Equal {Protection} {Under} the {Algorithm}: {A} {Legal}-{Inspired} {Framework} for {Identifying} {Discrimination} in {Machine} {Learning}},\n\tabstract = {Within the field of ethical machine learning, an area of special concern is the possibility of machine learning algorithms discriminating against groups of people in unethical ways, such as targeting advertisements based on race. In this paper, we propose a framework based on long-standing U.S. legal principles to determine whether the targeting of a group should be viewed with suspicion. Unlike existing work, we are focused on the case when the group is not correlated with known ‘protected features’, or such data is unavailable.},\n\tlanguage = {en},\n\tauthor = {Soundarajan, Sucheta and Clausen, Daniel L},\n\tyear = {2018},\n}\n\n","author_short":["Soundarajan, S.","Clausen, D. L"],"key":"soundarajan_equal_2018","id":"soundarajan_equal_2018","bibbaseid":"soundarajan-clausen-equalprotectionunderthealgorithmalegalinspiredframeworkforidentifyingdiscriminationinmachinelearning-2018","role":"author","urls":{},"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://bibbase.org/zotero/thomas_phsz","dataSources":["gDLCbJw3P8BMbaqME"],"keywords":[],"search_terms":["equal","protection","under","algorithm","legal","inspired","framework","identifying","discrimination","machine","learning","soundarajan","clausen"],"title":"Equal Protection Under the Algorithm: A Legal-Inspired Framework for Identifying Discrimination in Machine Learning","year":2018}