81,130 Hits in 6.7 sec

Post-processing of Differentially Private Data: A Fairness Perspective [article]

Keyu Zhu, Ferdinando Fioretto, Pascal Van Hentenryck
2022 arXiv   pre-print
Post-processing immunity is a fundamental property of differential privacy: it enables arbitrary data-independent transformations to differentially private outputs without affecting their privacy guarantees  ...  This paper shows that post-processing causes disparate impacts on individuals or groups and analyzes two critical settings: the release of differentially private datasets and the use of such private datasets  ...  The opinions expressed are solely those of the authors.  ... 
arXiv:2201.09425v1 fatcat:576mhksfenfldesvbidjnehwze

To Fair Value or Not to Fair Value: A Broader Perspective

Joshua Ronen
2008 Abacus. A Journal of Accounting and Business Studies  
private information.  ...  Fair value is considered here with respect to the two primary objectives of financial statements proposed in the joint conceptual framework that is under development by the FASB and the IASB, namely (a  ...  Yet these problems are mitigated by also providing objective data that place the expectations in proper perspective.  ... 
doi:10.1111/j.1467-6281.2008.00257.x fatcat:wvumgb2dgbee7bx55zcrnzf6ri

Learning Fair Representations for Recommendation: A Graph-based Perspective [article]

Le Wu, Lei Chen, Pengyang Shao, Richang Hong, Xiting Wang, Meng Wang
2021 arXiv   pre-print
Here, the fairness requirements refer to not exposing sensitive feature set in the user modeling process.  ...  In this paper, we propose a novel graph based technique for ensuring fairness of any recommendation models.  ...  ACKNOWLEDGEMENTS This work was supported in part by grants from the National Natural Science Foundation of China (Grant No. 61972125, U19A2079, U1936219, 61932009, 91846201), and CAAI-Huawei MindSpore  ... 
arXiv:2102.09140v3 fatcat:eul7bvqyrzbjjb5z7yg2p5d2vu

A Lawyer's Perspective: Guarantees of a Fair Trial and Online Dispute Resolution [chapter]

Nuno Ferreira Lousa
2022 Online Dispute Resolution  
where over the past 30 years he has taught Private International Law, Comparative Law, Civil Law and Intellectual Property Law.  ...  Member of the ICSID panels of arbitrators and conciliators and of the arbitration rosters of several arbitration centres in Angola, Brazil, Cape Verde, Mozambique and Portugal.  ...  be appropriate and would not compromise the rights of any party to a fair process.  ... 
doi:10.5771/9783748931508-59 fatcat:fjyw2kzltnbazd4wsvh5akosum

Causes and Consequences of the Student Protests in Chile [chapter]

Óscar Espinoza, Luis Eduardo González
2013 Fairness in Access to Higher Education in a Global Perspective  
state agency in charge of the entire accreditation system of tertiary institutions, as well as undergraduate and graduate programs; and (e) introducing an employment data collection observatory to provide  ...  From the evidence reviewed, this chapter concludes with a set of recommendations, which, if implemented, would contribute to an end of the crisis in Chile's tertiary system.  ...  The first is that the post-secondary studies generate a positive private rate of return therefore it should be financed by the beneficiaries who receive it.  ... 
doi:10.1007/978-94-6209-230-3_14 fatcat:urpiyidaevhxtj6wwylgsd326i

The Potential to Secure a Fair Trial Through Evidence Exclusion: A Swiss Perspective [chapter]

Laura Macula
2019 Ius Gentium: Comparative Perspectives on Law and Justice  
Swiss criminal procedure has a strong inquisitorial tradition and its primary purpose is the search for the "material truth.  ...  from the police perspective.  ...  Fair Trial The general principle of a fair trial, both in Swiss law and elsewhere, finds its origin in a variety of legal principles.  ... 
doi:10.1007/978-3-030-12520-2_2 fatcat:sbqcvykzajgzhof6v2okq465ta

Synthetic Data – what, why and how? [article]

James Jordon, Lukasz Szpruch, Florimond Houssiau, Mirko Bottarelli, Giovanni Cherubin, Carsten Maple, Samuel N. Cohen, Adrian Weller
2022 arXiv   pre-print
This explainer document aims to provide an overview of the current state of the rapidly expanding work on synthetic data technologies, with a particular focus on privacy.  ...  This article is intended to enable the reader to quickly become familiar with the notion of synthetic data, as well as understand some of the subtle intricacies that come with it.  ...  Composability and the post-processing theorem, in particular, allow for differentially private algorithms to be built out of smaller building blocks, which is precisely the driving force behind Differentially  ... 
arXiv:2205.03257v1 fatcat:2sdjqyfdivb4hnxn3jr2j5e6q4

Ethics of Open Data [article]

Nic Weber, Brandon Locke
2022 arXiv   pre-print
We begin with a brief overview of what can be thought of as three basic theories of ethics that intersect with dilemmas in openness, accountability, transparency, and fairness in data: Virtue, Consequential  ...  Throughout, we attempt to offer concrete examples of potential ethical dilemmas facing scholarly communication with respect to open data, and try to make clear what kinds of ethical positions are helpful  ...  From a Facebook user's perspective -the purposeful manipulation of how posts are promoted and displayed to their social network violates assumed virtue of the platform; users have a common and justified  ... 
arXiv:2205.10402v1 fatcat:plkfemfbdne4rg72tl7jr67gbu

Social Media Data - Our Ethical Conundrum

Lisa Singh, Agoritsa Polyzou, Yanchen Wang, Jason Farr, Carole Roan Gresenz
2020 IEEE Data Engineering Bulletin  
We frame the ethical dilemmas within the context of data privacy and algorithmic bias (fairness) and show how and when different ethical concerns arise.  ...  A recurrent theme throughout is the lack of transparency.  ...  Acknowledgements This research is funded by National Science Foundation awards #1934925 and #1934494, the National Collaborative on Gun Violence Research (NCGVR), the Fritz Family Fellowship, and the Massive Data  ... 
dblp:journals/debu/SinghPWFG20 fatcat:pirhmaonxfhzvmeawsxkx3hwua

Data Science Data Governance [AI Ethics]

Joshua A. Kroll
2018 IEEE Security and Privacy  
The shift to data-driven systems exacerbates gaps between traditional governance and oversight processes and the realities of software-driven decision making.  ...  How was a set of data collected or obtained? Were the data collected directly from customers, or was the dataset purchased from a third party?  ...  sample, such as when data come from a source of convenience (say, photo datasets culled from social network posts, which represent only the subset of users who post photos and could leave out or skew  ... 
doi:10.1109/msec.2018.2875329 fatcat:dwfmq4h43fecjirb4wp2ibmbei

The Data-Laundromat?

T. Gottschalk
2020 European Data Protection Law Review  
Based on the outcomes, the author proposes a change in the LED and discusses a centralised institution to govern access to novel data driven technology.  ...  To do so, the author searches for common ground of 'public availability' and reviews currently used technical approaches and common processing constellations.  ...  As a consequence, private data provides usually classify their processing as lawful speciously deemed lawful from a GDPR perspective.  ... 
doi:10.21552/edpl/2020/1/6 fatcat:fujzowg7hbcjrputfbz6c4qf5u

Fair Learning with Private Demographic Data [article]

Hussein Mozannar, Mesrob I. Ohannessian, Nathan Srebro
2020 arXiv   pre-print
Finally, we highlight how the methodology could apply to learning fair predictors in settings where protected attributes are only available for a subset of the data.  ...  We give a scheme that allows individuals to release their sensitive information privately while still allowing any downstream entity to learn non-discriminatory predictors.  ...  data is differentially private.  ... 
arXiv:2002.11651v2 fatcat:pca5cyxhdzcszm5lnopw6vvdmu

Generating private data with user customization [article]

Xiao Chen, Thomas Navidi, Ram Rajagopal
2020 arXiv   pre-print
Rather than training a large model to achieve privatization from end to end, we first decouple the creation of a latent representation, and then privatize the data that allows user-specific privatization  ...  We leverage a Variational Autoencoder (VAE) to create a compact latent representation of the data that remains fixed for all devices and all possible private labels.  ...  perspective.  ... 
arXiv:2012.01467v1 fatcat:pufrc6qx2feinbhrx2b2gjsm44

Statistical Data Privacy: A Song of Privacy and Utility [article]

Aleksandra Slavkovic, Jeremy Seeman
2022 arXiv   pre-print
Two dominant frameworks exist: statistical disclosure control (SDC), and more recent, differential privacy (DP).  ...  To quantify trade-offs between increasing demand for open data sharing and concerns about sensitive information disclosure, statistical data privacy (SDP) methodology analyzes data release mechanisms which  ...  The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.  ... 
arXiv:2205.03336v1 fatcat:r5dz3jb52rh7ji22r3a4erhdd4

Differential Data Protection Regimes in Data-Driven Research: Why the GDPR is More Research-Friendly Than You Think

Giovanni Comandè, Giulia Schneider
2022 German Law Journal  
AbstractAgainst the backdrop of an evolving landscape describing data driven research, this article discusses the role of data protection laws in shaping a free flow of research data.  ...  and of private data by public institutions, for either public interest-related research purposes or commercially oriented innovation purposes.  ...  The Notion of Research Under Data Protection Laws: A Comparative Perspective Big Data is deeply changing research methodology, and with it the range of public and private applications of the new insights  ... 
doi:10.1017/glj.2022.30 fatcat:wjhsimzmiraehi6bds32pbdtaa
« Previous Showing results 1 — 15 out of 81,130 results