It's not the algorithm, it's the data

@article{Kirkpatrick2017ItsNT,
  title={It's not the algorithm, it's the data},
  author={K. Kirkpatrick},
  journal={Communications of the ACM},
  year={2017},
  volume={60},
  pages={21 - 23}
}
In risk assessment and predictive policing, biased data can yield biased results. 

Topics from this paper

Identifying and Mitigating Bias in Machine Learning Applications
TLDR
There may be considerable ethical implications caused by machine learning bias; identifying and mitigating these biases is key to the development and deployment of effective machine learning algorithms. Expand
The Data Firehose and AI in Government: Why Data Management is a Key to Value and Ethics
TLDR
It is argued that data management capabilities are foundational to data analysis of any kind, but even more important in the present AI context, and that realizing the potential of AI for social good relies on investments specifically focused on this social outcome. Expand
Attaining Security Through Algorithms: Perspectives of Refugees and Data Experts
Originality/value: The study is based on a double security paradox where refugees who escape their homelands due to security concerns are also considered to be national security threats. DXs, on theExpand
MobilityMirror: Bias-Adjusted Transportation Datasets
TLDR
Customized synthetic datasets for publishing mobility data are described, showing how demographic biases arising from how the services are delivered may be amplified if released data is used in other contexts. Expand
Addressing Under-Reporting to Enhance Fairness and Accuracy in Mobility-based Crime Prediction
TLDR
A novel Bayesian Hierarchical model is proposed that utilizes domain knowledge about biases in reported crime data to characterize and enhance fairness and accuracy in mobility-based crime predictions. Expand
Transparency, Fairness, Data Protection, Neutrality
TLDR
Three recent regulatory frameworks are discussed: the European Union’s General Data Protection Regulation, the New York City Automated Decisions Systems (ADS) Law, and the Net Neutrality principle, which aim to protect the rights of individuals who are impacted by data collection and analysis. Expand
Transparency, Fairness, Data Protection, Neutrality: Data Management Challenges in the Face of New Regulation.
TLDR
Three recent regulatory frameworks that aim to protect the rights of individuals who are impacted by data collection and analysis are discussed: the European Union's General Data Protection Regulation, the New York City Automated Decisions Systems Law, and the Net Neutrality principle. Expand
Responsible data management
TLDR
It is argued that the data management community is uniquely positioned to lead the responsible design, development, use, and oversight of Automated Decision Systems. Expand
When Worlds Collide: Integrating Different Counterfactual Assumptions in Fairness
TLDR
This paper shows how it is possible to make predictions that are approximately fair with respect to multiple possible causal models at once, thus mitigating the problem of exact causal specification. Expand
FairPrep: Promoting Data to a First-Class Citizen in Studies on Fairness-Enhancing Interventions
TLDR
FairPrep is a design and evaluation framework for fairness-enhancing interventions based on a developer-centered design, and helps data scientists follow best practices in software engineering and machine learning, and shows how it can be used to measure the impact of sound best practices, such as hyperparameter tuning and feature scaling. Expand
...
1
2
3
...

References

New York State COMPAS-Probation Risk and Need Assessment Study: Examining the Recidivism Scale's Effectiveness and Predictive Accuracy
This report presents findings from a study which examined the effectiveness and predictive accuracy of the New York State COMPAS-Probation Recidivism Scale. This scale predicts the likelihood ofExpand