- Ethics and Social Impacts of AI
- Software Engineering Research
- Blockchain Technology Applications and Security
- Artificial Intelligence in Healthcare and Education
- Anomaly Detection Techniques and Applications
- Adversarial Robustness in Machine Learning
- Topic Modeling
- Explainable Artificial Intelligence (XAI)
- Machine Learning in Healthcare
Google (United States)
2020
Rising concern for the societal implications of artificial intelligence systems has inspired a wave academic and journalistic literature in which deployed are audited harm by investigators from outside organizations deploying algorithms. However, it remains challenging practitioners to identify harmful repercussions their own prior deployment, and, once deployed, emergent issues can become difficult or impossible trace back source.
Rising concern for the societal implications of artificial intelligence systems has inspired a wave academic and journalistic literature in which deployed are audited harm by investigators from outside organizations deploying algorithms. However, it remains challenging practitioners to identify harmful repercussions their own prior deployment, and, once deployed, emergent issues can become difficult or impossible trace back source. In this paper, we introduce framework algorithmic auditing...
Large technology firms face the problem of moderating content on their online platforms for compliance with laws and policies. To accomplish this at scale billions pieces per day, a combination human machine review are necessary to label content. Subjective judgement bias concern both annotated as well auditors who may be employed evaluate quality such annotations in conformance law and/or policy. address concern, paper presents novel application statistical analysis methods identify error...
This paper demonstrates how the limitations of pre-trained models and open evaluation datasets factor into assessing performance binary semantic similarity classification tasks. As (1) end-user-facing documentation around curation these model training regimes is often not easily accessible (2) given lower friction higher demand to quickly deploy such systems in real-world contexts, our study reinforces prior work showing disparities across datasets, embedding techniques distance metrics,...