Aydin Abadi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Aydin Abadi has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Marwan Zeggari
Lyzis Labs USA
2 shared publications - Mohamad Kassab
Boston University
2 shared publications - Renaud Lambiotte
Jiangsu University
2 shared publications - Steven J. Murdoch
University College London, London, United Kingdom
2 shared publications Aggelos Kiayias
University of Edinburgh
1 shared publicationBradley Doyle
1 shared publication
Latest Publications
- Open Access
Earn While You Reveal: Private Set Intersection that Rewards Participants
In Private Set Intersection protocols (PSIs), a non-empty result always reveals something about the private input sets of the parties. Moreover, in various variants of PSI, not all parties necessarily receive or are interested in the result. Nevertheless, to date, the literature has assumed that those parties who do not receive or are not interested in the result still contribute their private input sets to the PSI for free, although doing so would cost them their privacy. In this work, for the first time, we propose a multi-party PSI, called "Anesidora", that rewards parties who contribute their private input sets to the protocol. Anesidora is efficient; it mainly relies on symmetric key primitives and its computation and communication complexities are linear with the number of parties and set cardinality. It remains secure even if the majority of parties are corrupted by active colluding adversaries.
Open AccessEarn While You Reveal: Private Set Intersection that Rewards Participants
In Private Set Intersection protocols (PSIs), a non-empty result always reveals something about the private input sets of the parties. Moreover, in various variants of PSI, not all parties necessarily receive or are interested in the result. Nevertheless, to date, the literature has assumed that those parties who do not receive or are not interested in the result still contribute their private input sets to the PSI for free, although doing so would cost them their privacy. In this work, for the first time, we propose a multi-party PSI, called "Anesidora", that rewards parties who contribute their private input sets to the protocol. Anesidora is efficient; it mainly relies on symmetric key primitives and its computation and communication complexities are linear with the number of parties and set cardinality. It remains secure even if the majority of parties are corrupted by active colluding adversaries.
- Open Access
Starlit: Privacy-Preserving Federated Learning to Enhance Financial Fraud Detection
Federated Learning (FL) is a data-minimization approach enabling collaborative model training across diverse clients with local data, avoiding direct data exchange. However, state-of-the-art FL solutions to identify fraudulent financial transactions exhibit a subset of the following limitations. They (1) lack a formal security definition and proof, (2) assume prior freezing of suspicious customers' accounts by financial institutions (limiting the solutions' adoption), (3) scale poorly, involving either $O(n^2)$ computationally expensive modular exponentiation (where $n$ is the total number of financial institutions) or highly inefficient fully homomorphic encryption, (4) assume the parties have already completed the identity alignment phase, hence excluding it from the implementation, performance evaluation, and security analysis, and (5) struggle to resist clients' dropouts. This work introduces Starlit, a novel scalable privacy-preserving FL mechanism that overcomes these limitations. It has various applications, such as enhancing financial fraud detection, mitigating terrorism, and enhancing digital health. We implemented Starlit and conducted a thorough performance analysis using synthetic data from a key player in global financial transactions. The evaluation indicates Starlit's scalability, efficiency, and accuracy.Open AccessStarlit: Privacy-Preserving Federated Learning to Enhance Financial Fraud Detection
Federated Learning (FL) is a data-minimization approach enabling collaborative model training across diverse clients with local data, avoiding direct data exchange. However, state-of-the-art FL solutions to identify fraudulent financial transactions exhibit a subset of the following limitations. They (1) lack a formal security definition and proof, (2) assume prior freezing of suspicious customers' accounts by financial institutions (limiting the solutions' adoption), (3) scale poorly, involving either $O(n^2)$ computationally expensive modular exponentiation (where $n$ is the total number of financial institutions) or highly inefficient fully homomorphic encryption, (4) assume the parties have already completed the identity alignment phase, hence excluding it from the implementation, performance evaluation, and security analysis, and (5) struggle to resist clients' dropouts. This work introduces Starlit, a novel scalable privacy-preserving FL mechanism that overcomes these limitations. It has various applications, such as enhancing financial fraud detection, mitigating terrorism, and enhancing digital health. We implemented Starlit and conducted a thorough performance analysis using synthetic data from a key player in global financial transactions. The evaluation indicates Starlit's scalability, efficiency, and accuracy. - Open Access
Safeguarding Physical Sneaker Sale Through a Decentralized Medium
Sneakers were designated as the most counterfeited fashion item online, with three times more risk in a trade than any other fashion purchase. As the market expands, the current sneaker scene displays several vulnerabilities and trust flaws, mostly related to the legitimacy of assets or actors. In this paper, we investigate various blockchain-based mechanisms to address these large-scale trust issues. We argue that (i) pre-certified and tracked assets through the use of non-fungible tokens can ensure the genuine nature of an asset and authenticate its owner more effectively during peer-to-peer trading across a marketplace; (ii) a game-theoretic-based system with economic incentives for participating users can greatly reduce the rate of online fraud and address missed delivery deadlines; (iii) a decentralized dispute resolution system biased in favour of an honest party can solve potential conflicts more reliably.
Open AccessSafeguarding Physical Sneaker Sale Through a Decentralized Medium
Sneakers were designated as the most counterfeited fashion item online, with three times more risk in a trade than any other fashion purchase. As the market expands, the current sneaker scene displays several vulnerabilities and trust flaws, mostly related to the legitimacy of assets or actors. In this paper, we investigate various blockchain-based mechanisms to address these large-scale trust issues. We argue that (i) pre-certified and tracked assets through the use of non-fungible tokens can ensure the genuine nature of an asset and authenticate its owner more effectively during peer-to-peer trading across a marketplace; (ii) a game-theoretic-based system with economic incentives for participating users can greatly reduce the rate of online fraud and address missed delivery deadlines; (iii) a decentralized dispute resolution system biased in favour of an honest party can solve potential conflicts more reliably.
- Open Access
An Efficient and Decentralized Blockchain-based Commercial Alternative (Full Version)
While online interactions and exchanges have grown exponentially over the past decade, most commercial infrastructures still operate through centralized protocols, and their success essentially depends on trust between different economic actors. Digital advances such as blockchain technology has led to a massive wave of \textit{Decentralized Ledger Technology} (\textit{DLT}) initiatives, protocols and solutions. This advance makes it possible to implement trustless systems in the real world, which, combined with appropriate economic and participatory incentives, would foster the proper functioning and drive the adoption of a decentralized platform among different actors. This paper describes an alternative to current commercial structures and networks by introducing \textit{Lyzis Labs}, which is is an incentive-driven and democratic protocol designed to support a decentralized online marketplace, based on blockchain technology. The proposal, \textit{Lyzis Marketplace}, allows to connect two or more people in a decentralized and secure way without having to rely on a \textit{Trusted Third Party} (\textit{TTP}) in order to perform physical asset exchanges while mainly providing transparent and fully protected data storage. This approach can potentially lead to the creation of a permissionless, efficient, secure and transparent business environment where each user can gain purchasing and decision-making power by supporting the collective welfare while following their personal interests during their various interactions on the network.
Open AccessAn Efficient and Decentralized Blockchain-based Commercial Alternative (Full Version)
While online interactions and exchanges have grown exponentially over the past decade, most commercial infrastructures still operate through centralized protocols, and their success essentially depends on trust between different economic actors. Digital advances such as blockchain technology has led to a massive wave of \textit{Decentralized Ledger Technology} (\textit{DLT}) initiatives, protocols and solutions. This advance makes it possible to implement trustless systems in the real world, which, combined with appropriate economic and participatory incentives, would foster the proper functioning and drive the adoption of a decentralized platform among different actors. This paper describes an alternative to current commercial structures and networks by introducing \textit{Lyzis Labs}, which is is an incentive-driven and democratic protocol designed to support a decentralized online marketplace, based on blockchain technology. The proposal, \textit{Lyzis Marketplace}, allows to connect two or more people in a decentralized and secure way without having to rely on a \textit{Trusted Third Party} (\textit{TTP}) in order to perform physical asset exchanges while mainly providing transparent and fully protected data storage. This approach can potentially lead to the creation of a permissionless, efficient, secure and transparent business environment where each user can gain purchasing and decision-making power by supporting the collective welfare while following their personal interests during their various interactions on the network.
- 22 October 2022
- Published by Springer Nature
- p. 100-119
Multi-party Updatable Delegated Private Set Intersection
- 22 October 2022
- Published by Springer Nature
- p. 100-119
Multi-party Updatable Delegated Private Set Intersection
- 23 October 2021
- Published by Springer Nature
- p. 541-559
Multi-instance Publicly Verifiable Time-Lock Puzzle and Its Applications
- 23 October 2021
- Published by Springer Nature
- p. 541-559
Multi-instance Publicly Verifiable Time-Lock Puzzle and Its Applications
- 2 October 2021
- Published by Springer Nature
- p. 721-742
Polynomial Representation Is Tricky: Maliciously Secure Private Set Intersection Revisited
- 2 October 2021
- Published by Springer Nature
- p. 721-742
Polynomial Representation Is Tricky: Maliciously Secure Private Set Intersection Revisited