Fairness Score and process standardization: framework for fairness certification in artificial intelligence systems
نویسندگان
چکیده
Decisions made by various Artificial Intelligence (AI) systems greatly influence our day-to-day lives. With the increasing use of AI systems, it becomes crucial to know that they are fair, identify underlying biases in their decision-making, and create a standardized framework ascertain fairness. In this paper, we propose novel Fairness Score measure fairness data-driven system Standard Operating Procedure (SOP) for issuing Certification such systems. audit process standardization will ensure quality, reduce ambiguity, enable comparison improve trustworthiness It also provide operationalise concept facilitate commercial deployment Furthermore, Certificate issued designated third-party auditing agency following would boost conviction organizations intend deploy. The Bias Index proposed paper reveals comparative bias amongst protected attributes within dataset. To substantiate framework, iteratively train model on biased unbiased data using multiple datasets check correctly judge
منابع مشابه
test score equating and fairness in language assessment
equating test scores is an important issue in large scale testing. almost all standardized tests have several forms which vary in difficulty. new forms are written and added every year. when the items in different forms of a test vary in difficulty, direct comparison of test-takers who have taken different forms and are at the same ability level is not possible; hence, the issue of test fairnes...
متن کاملA continuous framework for fairness
Increasingly, discrimination by algorithms is perceived as a societal and legal problem. As a response, a number of criteria for implementing algorithmic fairness in machine learning have been developed in the literature. This paper proposes the Continuous Fairness Algorithm (CFA$\theta$) which enables a continuous interpolation between different fairness definitions. More specifically, we make...
متن کاملArtificial agents learning human fairness
Recent advances in technology allow multi-agent systems to be deployed in cooperation with or as a service for humans. Typically, those systems are designed assuming individually rational agents, according to the principles of classical game theory. However, research in the field of behavioral economics has shown that humans are not purely self-interested: they strongly care about fairness. The...
متن کاملFairness in Action Systems
An action system is a framework for describing parallel or distributed systems, for which the reenement calculus ooers a formalisation of the stepwise development method. Fairness is a useful notion in modelling parallel or distributed systems, and this paper extends the reenement calculus to deal with fair action systems. Simulations of fair action systems are obtained by extending simulations...
متن کاملFairness for Infinite-State Systems
In this paper we introduce the first known tool for symbolically proving fair -CTL properties of (infinite-state) integer programs. Our solution is based on a reduction to existing techniques for fairnessfree CTL model checking via the use of infinite non-deterministic branching to symbolically partition fair from unfair executions. We show the viability of our approach in practice using exampl...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: AI and ethics
سال: 2022
ISSN: ['2730-5953', '2730-5961']
DOI: https://doi.org/10.1007/s43681-022-00147-7