A5091625866- Ethics and Social Impacts of AI
- Psychology of Moral and Emotional Judgment
- Social and Intergroup Psychology
- Explainable Artificial Intelligence (XAI)
- Hate Speech and Cyberbullying Detection
- Data Stream Mining Techniques
- Transportation and Mobility Innovations
- Terrorism, Counterterrorism, and Political Violence
- Privacy, Security, and Data Protection
- Transportation Planning and Optimization
- Adversarial Robustness in Machine Learning
- Privacy-Preserving Technologies in Data
- Mobile Crowdsensing and Crowdsourcing
- Ethics in Business and Education
- Urban Transport and Accessibility
- Urban Green Space and Health
- Land Use and Ecosystem Services
- Energy, Environment, and Transportation Policies
- Opinion Dynamics and Social Influence
- Team Dynamics and Performance
- Digital Economy and Work Transformation
- Experimental Behavioral Economics Studies
- Smart Grid Energy Management
- Game Theory and Voting Systems
- Social Power and Status Dynamics
University of Southern California
2019-2024
Integrated Systems Incorporation (United States)
2019-2021
With the widespread use of AI systems and applications in our everyday lives, it is important to take fairness issues into consideration while designing engineering these types systems. Such can be used many sensitive environments make life-changing decisions; thus, crucial ensure that decisions do not reflect discriminatory behavior toward certain groups or populations. We have recently seen work machine learning, natural language processing, deep learning addresses such challenges...
What is the best way to define algorithmic fairness? While many definitions of fairness have been proposed in computer science literature, there no clear agreement over a particular definition. In this work, we investigate ordinary people's perceptions three these definitions. Across two online experiments, test which people perceive be fairest context loan decisions, and whether change with addition sensitive information (i.e., race applicants). Overall, one definition (calibrated fairness)...
What is the best way to define algorithmic fairness? While many definitions of fairness have been proposed in computer science literature, there no clear agreement over a particular definition. In this work, we investigate ordinary people's perceptions three these definitions. Across two online experiments, test which people perceive be fairest context loan decisions, and whether change with addition sensitive information (i.e., race applicants). Overall, one definition (calibrated fairness)...
Online radicalization is among the most vexing challenges world faces today. Here, we demonstrate that homogeneity in moral concerns results increased levels of radical intentions. In Study 1, find Gab—a right-wing extremist network—the degree convergence within a cluster predicts number hate-speech messages members post. 2, replicate this observation another network, Incels. Studies 3 to 5 ( N = 1,431), experimentally leading people believe others their hypothetical or real group share...
Bias in machine learning has rightly received significant attention over the past decade. However, most fair (fair-ML) works to address bias decision-making systems focused solely on offline setting. Despite wide prevalence of online real world, work identifying and correcting setting is severely lacking. The unique challenges environment make addressing more difficult than First, Streaming Machine Learning (SML) algorithms must deal with constantly evolving real-time data stream. Secondly,...
Abstract Ride-hailing services have skyrocketed in popularity due to their convenience. However, recent research has shown that pricing strategies can a disparate impact on some riders, such as those living disadvantaged neighborhoods with greater share of residents color or below the poverty line. Analyzing real-world data, we additionally show these communities tend be more dependent ride-hailing (e.g., for work commutes) lack adequate public transportation infrastructure. To this end,...
No abstract available.
Online radicalization is among the most vexing challenges world faces today. Here, we demonstrate that homogeneity in moral concerns results increased levels of radical intentions. In Study 1, find Gab—a right-wing extremist network—the degree convergence within a cluster predicts number hate-speech messages members post. 2, replicate this observation another network, Incels. Studies 3 to 5 (N = 1,431), experimentally leading people believe others their hypothetical or real group share views...
Despite location being increasingly used in decision-making systems employed many sensitive domains such as mortgages and insurance, astonishingly little attention has been paid to unfairness that may seep due the correlation of with characteristics considered protected under anti-discrimination law, race or national origin. This position paper argues for urgent need consider fairness respect location, termed \textit{spatial fairness}, by outlining harms continue be perpetuated location's...
Ride-hailing services have skyrocketed in popularity due to the convenience they offer, but recent research has shown that their pricing strategies can a disparate impact on some riders, such as those living disadvantaged neighborhoods with greater share of residents color or below poverty line. Since these communities tend be more dependent ride-hailing lack adequate public transportation, it is imperative address this inequity. To end, paper presents first thorough study fair for by...
Bias in machine learning has rightly received significant attention over the last decade. However, most fair (fair-ML) work to address bias decision-making systems focused solely on offline setting. Despite wide prevalence of online real world, identifying and correcting setting is severely lacking. The unique challenges environment make addressing more difficult than First, Streaming Machine Learning (SML) algorithms must deal with constantly evolving real-time data stream. Second, they...
Humans spend a significant part of their lives being groups. In this document we propose research directions that would make it possible to computationally form productive We bring light several issues need be addressed in the pursuit goal (not amplifying existing biases and inequality, for example), as well multiple avenues study help achieve us task efficiently.