From the course: Ethics and Law in Data Analytics

Best practices to remove bias

From the course: Ethics and Law in Data Analytics

Start my 1-month free trial

Best practices to remove bias

“

- In this video, we're going to look at some of the best practices suggested from the White House big data report of 2016 to eliminate bias and to prevent discrimination. So this quote is directly from the report itself. Moving forward, it's essential that the public and private sectors continue to have collaborative conversations about how to achieve the most out of big data technologies while deliberatively applying these tools to avoid, and when appropriate, address, discrimination. Best practice one. The White House suggests supporting research into to mitigating algorithmic discrimination, building systems that support fairness and accountability, and developing strong data ethics frameworks. So you see, the usefulness of this course is even more obvious in this first best practice. We're learning about ethics and legal frameworks to prevent discrimination and bias. This learning is essential not only because it's suggested as a best practice, but because will see as the law does eventually catch up to the technology, those of us who are prepared and understand ethics and legal frameworks will be in a strategic position to do better in this industry. The second best practice suggested by the White House is to encourage market participants to design the best systems including transparency and accountability mechanisms such as the ability for subjects to correct inaccurate data and to appeal algorithmic-based decisions. So this practice is coming out of what we know to be issues of the lack of transparency, the lack of knowing about what's happening behind the scenes and it's meant to encourage those that are in the industry to not wait for legal regulatory reform to be enforced upon them, but to take action and to use design practices to address the problems. The third practice, promoting academic research and industry development of algorithmic auditing and external testing of big data systems to ensure that people are being treated fairly. This is a little bit like the second big practice and I would say that Nathan has addressed both of these practices in his videos and his discussions about design systems and the importance on continually researching and developing in this area. The fourth one, broaden participation in computer science and data science, including opportunities to improve basic fluencies and capabilities of all Americans. Super important here. So many people are just not really aware at all about what is happening in the big data revolution, that this data's being collected, that it's being used, that there's possibilities of bias and discrimination. They don't know to advocate for themselves, they don't understand the science. So I would say that those of you participating in this course today are ahead of the game, that you're taking a course that been developed by Microsoft to actually broaden the participation in computer science and data science so that we're learning about this so that we can become better professionals. Universities like Seattle University are also taking a lead in this by designing graduate programs and undergraduate programs in data analytics and law and ethics so that in learning about the science, we're also learning about the ethical and legal frameworks that are important. The fifth best practice, consider the role of the government and private sector in setting the rules of the road for how data is used. I would also say here that you also have to consider the role of civil society. There are a lot of nonprofits, think tanks, and groups that are involved in understanding what is happening in data analytics and artificial intelligence and those voices need to be heard as well.

Contents