Software company dedicated to providing payment solutions to influencer marketing agencies and creators’ businesses

Zexel

HyperTransformer: Conclusion and References

Transductive Learning for Textual Few-Shot: Limitations, Acknowledgements, & References

Verifiable Privacy-Preserving Computing: Conclusion & References

This paper is under CC 4.0 license. available on arxiv Authors: (1) Tariq Bontekoe, University of Groningen; (2) Dimka Karastoyanova, University of Groningen; (3) Fatih Turkmen, University of Groningen. Table of Links Abstract & Introduction Preliminaries Zero-knowledge proofs and verifiable computing Privacy-preserving computations Requirements: an application’s perspective Verifiable, privacy-preserving computing Open challenges and future directions Conclusion & References 5. Requirements: an application’s perspective In this section, we discuss how verifiability is combined with privacy-preserving computing in practical settings that are frequently observed for verifiability and/or privacypreserving computations. We discuss two frequent scenarios: verifiable outsourcing and blockchain; and two use cases showing specific requirements: e-voting and genomics. These were the most frequently observed scenarios and use cases in existing literature. For each application we summarize the domain-specific requirements and efficiency trade-offs, that answer when combined. Each property has a specific mark X assigned, that will be used for reference in subsequent sections. RQ2 5.1. Verifiable outsourcing One of the first and most straightforward applications of verifiable PETs is that of verifiable outsourcing, where a (group of) data owner(s) outsource(s) a costly computation to a computing party with more resources. This computing party produces a computation result as well as a ZKP, trusted execution environment (TEE) attestation, or MAC to guarantee correctness. When the used verification method allows for verification that is more efficient than the original computation, the scheme is considered efficient. Pinocchio [21] was one of the first efficient ZKP schemes for general computations, making it practical for verifiable outsourcing. Since then, many new ZKPs schemes with better efficiency or security have been constructed, e.g., [22], [25], [49], [50]. However, when applying these schemes directly, they do not provide data privacy in most use cases, even though this is a necessity when computing over private data distributed over multiple parties. Since, in most cases, only the clients that outsource their computation need to be able to guarantee correctness, public verifiability is often not necessary. Most importantly, verifiable outsourcing is only beneficial when the verification (and/or preprocessing) time of the clients is significantly smaller than the time it would take to compute the outcome by themselves, i.e., the solution should be efficient. 5.2. Blockchain Blockchain was introduced as a platform to realize decentralized online currency, but has since become a much broader platform. The recent surge in blockchain and DLT applications, such as dApps, has also had significant impact on the development of new ZKP schemes, an important solution for verifiability. However, blockchain applications also frequently rely upon ZKPs to provide applications that guarantee correctness, whilst maintaining privacy. Examples of such applications range from private monetary transactions, e.g., Zerocash [7], Zether [26], and others [51], [52], to private smart contracts, e.g., Hawk [53], and decentralized private computation (DPC), e.g., ZEXE [54]. Due to the decentralized nature of blockchain, with a lack of central authority to guarantee correctness over private data, verifiable computing methods are a logical, and often unavoidable, solution. For most blockchain applications, regular verifiability will often not be sufficient. Since every (future) connected party must be able to verify any message posted to the blockchain, public verifiability is a necessary requirement. To prevent scaling issues in blockchain applications, developers attempt to keep verification times low, and onchain messages as small as possible. Therefore, solutions that sacrifice prover time for lower verification times and sublinear (or constant) sized verification messages are favorable. Finally, we note that interactive protocols are, generally speaking, not suitable for distributed ledgers as that would require direct communication, conflicting the principles of DLT. Since non-interactivity is a prerequisite for public verifiability, we do not include this requirement by itself. 5.3. Verifiable privacy-preserving genomics Usage of privacy-preserving computations in genomics has taken a big leap in the past decade [55], [56], [57] and even led to a yearly competition on human genome privacy [58], [59], [60], with specific tracks on the use of PETs and blockchain. The first application of verifiable computing in the genomic domain is discussed in [61], as far as we are aware. Specifically, the authors propose the use of HE and homomorphic MACs to outsource computation of the disease susceptibility (DS) score to a third party in a privacypreserving and verifiable manner. The increase in efficient methods for verifiable computing in recent years, has also led to the first solutions in this domain. A method to perform an elastic net regression among multiple parties in a privacy-preserving manner, by outsourcing the computation to a pair of non-colluding cloud providers, is proposed in [62]. Correctness of the result is guaranteed by verifying that the evaluation of the final gradient is close to zero. [63] presents a solution for evaluating secure count or top-k queries on multi-source genomic data. All data is encrypted using an additively homomorphic encryption scheme and shared with two non-colluding cloud providers. The cloud parties evaluate the query, with privacy of query, input data, and result, and use MACs to guarantee correctness. [64] presents a verifiable homomorphic secret sharing scheme for evaluating polynomials of low degree. Verifiable genomic analyses, and verifiable computations on outsourced data are mentioned as possible use cases. Lastly, [65] computes the minor allele frequency (MAF) score over multi-source data in a blockchain setting. The more expensive computations, those on additively HE ciphertexts, are performed off-chain, ZKPs are used to guarantee correctness thereof. It is clear that the use cases in genomics are very diverse and have different requirements regarding efficiency or which parties can verify. However, it is abundantly clear that genomic data requires strong privacy guarantees. Due to the long-livedness of genomic data, i.e., a person’s genome also reveals significant private information on their family and descendants, store-now-decrypt-later attacks form a serious risk, especially at the dawn of practical quantum computing [66]. The use of post-quantum secure solutions should therefore be seriously considered for any work in the genomic domain. 5.4. Verifiable e-voting An electronic voting system allows a population to cast their votes using electronic rather than physical means. The need for data privacy follows directly from the fact that, in general, elections and voting processes require voter privacy. Using verifiability techniques to make elections completely verifiable by the general public has been a topic of active research for many years, e.g., [67], [68]. The developments in blockchain and more efficient ZKPs have led to many new proposals for e-voting solutions. [69] discusses the use of NIZK proofs to make threshold homomorphic encryption-based voting secure. A novel SNARK-friendly, additively-homomorphic verifiable encryption scheme is proposed and applied to verify vote validity is presented in [9]. Recently, [70] proposes an end-to-end verifiable and secret election scheme, based on NIZKs and MPC. Blockchains and cloud servers can both be used to store the secure ballots. It is easy to see that public verifiability is a requirement in verifiable e-voting schemes. Next to this, trust is one of the most important factors to guarantee that e-voting schemes are to be adopted in practice [71]. It may therefore make sense, to only use solutions that are post-quantum secure and rely on standard assumptions to improve the level of trust, potentially at the cost of some efficiency.

Applications of Verifiable Privacy-Preserving Computing

Preface: This is a comic that contains artwork that is largely created using a generative AI under my direction. Hope you all enjoy the story. Thoth is known as the "Ice World" for its stable sub-arctic cold conditions. Jian Xi is a researcher in stasis biology, i.e. suspended animation. He serves The Alliance at the station on Thoth, known as the "Ice World" for its stable sub-arctic cold conditions. Water and ice, mostly ice, cover everything. The Alliance is here because of our interest in these guys, the Zeels. They eat fish and flourish in the cold ecosystem. And this guy, who also eats fish, but larger fish. Of course some of the fish also eat fish. And so do we. There are two realtively new research facilities here at Swerdlov Base. This structure houses operations that are basically doing pure research. This facility is a bit newer and besides research most of our production work is also done here We have a number of these snocat tractors that serve for basic transportation though the terrain is too inhospitable even for them at any distance from the base. The original Sverdlov Base was this structure which still houses our power generation facility. The base was originally set up by The Hegemony, but now The Alliance controls everything. Most of our work involves these stasis modules where volunteers basically enter a state of suspended animation wherein they are pretty much frozen in time. In fact we have some of these guys who have been chilling for a LONG time, decades or even centuries in a few cases. The Alliance figures this might be a method to reach destinations otherwise beyond our grasp. Part of our research is focused on their dreams, we are testing "the bug juice" from Shanzi. All very interesting in an arcane sort of a way. Meanwhile, when we revive them or remotivate them, the view will be much the same as it was then they went under. Cold and colder. I hope you enjoyed this scifi comic novel. Sequels of a sort relating more about The Alliance follow. As always, Comments, Criticisms, and Suggestions are Welcome. Thank you for reading!

Thoth, The Ice World

TO PROVE BY AUTHORITY A PASSAGE TO BE ON THE NORTH SIDE OF AMERICA

References
Zcash - Aleo - Zcash protocol specification Zexe protocol specification Zcash 1. About Zcash ? A to learn about Zcash. short video Features: Anonymous version of BTC, UTXO-like model Can only be used for payment scenarios with no programmability 2. Main Concepts Note: Zcash has undergone several and we only focus on the latest version. This article mainly introduces the core concepts of Zcash. protocol upgrades 2.1 Key Components sk : spend key, generated uniformly ask: spend authorizing key ak : spend validating key nk : nullifier deriving key rivk : randomness dk: diversifier key ivk : private key / incoming viewing key ovk : outgoing viewing key pkd : transmission key Spending key: secret key Full viewing key: decrypt the transaction context Incoming viewing key Outgoing viewing key Shielded payment address: should be different each time You can learn about the computation methods of thesekeys in the Zcash protocol specification: section 4.2.3, page 36. 2.2 Notes Note is the basic unit of the Zcash protocol, similar to the UTXO of BTC; in Zcash, the inputs and outputs of all transactions are notes. Of course, Zcash also supports non-anonymous transactions, identical to Bitcoin’s transaction model. Therefore, in order to further understand Zcash, you must know the data structure of note first: {d,pk_d}: address information of the owner of the note v : the corresponding amount of the note {ρ,ψ}: compute the random number of the nullifier rcm : being used to compute the random number of the note commitment In the Zcash protocol, the note cannot be made public due to the requirement of privacy. Therefore, the corresponding commitment needs to be computed to represent the note. The computation method is as follows: 2.3 Action Transfer A transaction may contain multiple action transfers, and each action transfer will spend the old note and generate a new note. The data structure is as follows: cv^{net}: being used for the balance check for action transfer in the form of binding signature without being included in the zk proof rt^{Orchard}: the treestate of the previous block, which is used to verify the spent note validity nf: the logo of the spent note, which is used to prevent the double spend rk: being used to verify that the spender has the right to the spend the note in the form of spendAuthSig signature verification cm_x: commitment of generating a new note epkepk temporary public key, which is used to decrypt the noteplaint information of the new note C^{enc},C^{out}: the encrypted ciphertext of the new note enbaleSpends, enableOutputs: indicating the type of the current action transfer π: zk proof 2.4 Action Statement The public inputs are: The privacy inputs are: The statements to prove: Spent note integrity is uniquely bound to the noteplaint Spent note validity, existence proof of the cm tree Value commitment integrity is uniquely bound to rcv , old value, and new value Nullifier integrity prevents double spend and maintains a spent note set Spent note legitimacy Address integrity New note integrity Flag legitimacy 2.5 Transaction Constructions and Examples 2.5.1 Transaction construction The entire transaction structure consists of four parts: Public info (1 - 5) Transparent transactions info (6 - 9) Sapling transactions info (10 - 16) Orchard transaction info (17 - 25) 2.5.2 From Transparent to Shield The Orchard protocol includes two types of addresses, transparent address (TA) and shield address (SA). Generally, in order to execute a private transaction, it is necessary to transfer from TA to SA first. The corresponding transaction structure should be: Public info (1 - 5) Transparent transactions info (6 - 9) tx_in _* : actual value tx_out _* : default value Sapling transactions info (10 - 16) All : default value Orchard transaction info (17 - 25) All: actual value 2.5.3 From Shield to Shield The Orchard protocol includes two types of addresses, transparent address (TA) and shield address (SA). Generally, in order to execute a private transaction, it is necessary to transfer from TA to SA first. The corresponding transaction structure should be: Public info (1 - 5) Transparent transactions info (6 - 9) All : default value Sapling transactions info (10 - 16) All : default value Orchard transaction info (17 - 25) All: actual value 2.5.4 From Shield to Transparent The Orchard protocol includes two types of addresses, transparent address (TA) and shield address (SA). Generally, in order to execute a private transaction, it is necessary to transfer from TA to SA first. The corresponding transaction structure should be: Public info (1 - 5) Transparent transactions info (6 - 9) tx_in _* : default value tx_out _* : actual value Sapling transactions info (10 - 16) All : default value Orchard transaction info (17 - 25) All: actual value 2.6 How is Privacy Achieved? The generated note is represented by cm, and the spent note is represented by nf. There is no connection between nf and cm. Therefore, no one can know in which transaction any generated note was spent. Unlinkable Private The transaction information does not contain the sender address and the spendAuthSig is a one-time signature (it is different every time, so the public key is different, rk) Sender address: The transaction does not contain the receiver’s address and the new note plaint is encrypted by the receiver’s public key (the receiver’s private address is also one-time) Receiver address: Hide the note in the form of pedersen commitment, and ensure the balance attribute of the transaction by bindsig Value: Aleo 1. The Similarities and Differences With Zcash Zcash can only execute private transactions based on the OUTX model and have no programmability; therefore, the main difference between Aleo and Zcash is privacy programmability; and the similarity is that they both support privacy attributes (transaction privacy is not only limited to assets). 2. Aleo vs Zcash 2.1 Unit Unlike the note of Zcash, the basic operation unit of Aleo is record (UTXO in BTC). Find the main differences between the two below: Although the names of the specific parameters are different, there is a corresponding relationship between the two from the functional point of view: corresponding to the address information of the note owner, commitment-related information, nf/sn-related information, and value-related information, respectively. Therefore, the structures of the two are quite similar; the main differences exist in the birth predicate and death predicate of the record, which are two Boolean-type functions, representing the conditions that need to be met when a record is in the birth(generate) and death(spend) stages. This part supports user-defined, so it’s programmable. 2.2 Transaction Construction There are still some resemblances compared with the main construction of the transaction of Zcash (2.5.1): The corresponding serial number sn of the spent record, which is represented by nf in Zcash, is unique The commitment of the new record The plaint of the new record, including owner information, corresponding birth/death predicate, etc 2.3 Prover Statement You need to prove: The old record validity The old record legitimacy (have the right to spend the record) The new record validity The birth/death predicate validity (similar to the balance check in Zcash) 3. More 3.1 Technologies not mentioned From the perspective of the paper, the privacy design of Aleo 's programmable privacy design is more similar to the earlier Zcash white paper ( ), the similar key structure, the similar note structure and the similar name (nf is called sn in zerocash, serial number). This article makes a comparison based on the latest paper of Zcash and the ZEXE of Aleo. Although there are differences in specific details, such as the key structure and the specific cryptographic methods used, the high-level design is generally the same. In addition to the technical details described above, there are still some other technical details that have not been mentioned yet, such as the delegate prover scheme, zero-knowledge proof algorithm, recursion/aggregation scheme, and so on. People interested in them can study further. zerocash 3.2 Why are they all UTXO -based, not account-based? Remark 2.3 (Zexe protocol specification: section 2.3, page 11)

Private Transactions: Insights to the Design Principles in ZCash and Aleo

Transacciones privadas: información sobre los principios de diseño en ZCash y Aleo

Everyone wants to go Agile today. Teams want to put the user in the centre of their product development process while building products. After all, you are building the product for your users, right? To do this, teams write user stories as a way to keep the user in mind while defining the product and its features. And to ensure the team meets the standards their users expect, teams write acceptance criteria for each user story. It is easy for teams to neglect acceptance criteria as being simple. However, they're used to document customer expectations, provide an end-user perspective, clarify requirements and prevent ambiguity. When done properly, this will eventually verify if the development goals were met. help your quality assurance In this article I will talk about some of the common mistakes teams make when writing user stories. 1. They are too broad Writing acceptance criteria too broadly defeats the fundamental purpose of its existence. While they should not be confused with test cases, acceptance criteria should communicate what specific conditions should be met to ensure that when a is completed, it meets a certain standard. user story 2. Does not consider the user experience A good product is table stakes today. And that doesn't mean spewing features one after another, but building products and features that provides good experiences at every touch point. This makes user experience more important than what it was before. If your acceptance criteria doesn't take into consideration the end-user experience, you're most likely missing out on a huge opportunity to satisfy your customers. 3. Writing about the "how" instead of the "what" Yes, acceptance criteria should be specific and concise. But that shouldn't make you get too specific that you end up talking about the implementation details of how you're going to go about it. Remember, at the end of the day, acceptance criteria are a set of criteria that needs to be satisfied. If they include "how" your team should implement a user story, you are in a way micromanaging them. By keeping things simple and talking about the "what", you give your team the opportunity to explore, get creative, and come up with solutions to solve the problem at hand. 4. Writing acceptance criteria as you build Although acceptance criteria is usually written by the product owner, it can be written by anyone in the cross-functional team. Sometimes, I've noticed dev teams take this a bit too far by writing acceptance criteria while they're building the feature. The problem with this is, it doesn't keep other members aligned on this criteria and can lead to mismatch in expectations. It's essential for teams to ensure they write acceptance criteria before the sprint begins — ideally before the . sprint planning meeting 5. Having assumptions It can feel silly to write an acceptance criteria that you think is kind of obvious. But what's obvious for you, need not necessarily make it obvious for your team. Building products and features is a cross-functional effort. That means, people come with different perspectives that others don't know. Writing even the most obvious acceptance criteria can help your team give the complete picture of what needs to be done and help remove inefficiencies in your development process. Conclusion Your might help you writing acceptance criteria with a template. But it is never as simple as completing a fill-in-the-blanks form. The individual impact of inefficiencies in your product development process is almost always invisible. One mistake while writing an acceptance criteria often leads to a series of other mistakes. And the worst part is, it is nearly impossible to trace it back to a poorly written acceptance criteria. agile tool Reference https://zepel.io/agile/user-stories/acceptance-criteria/

Zexel

zexel.io

-- employees

Since n.d.

COMPANY RANKING

ZEXEL

EVERGREEN INDEX #12083

Zexel WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

Zexel

zexel.io

-- employees

Since n.d.

COMPANY RANKING

ZEXEL

EVERGREEN INDEX #12083

RELATED COMPANIESRANK

STORIES ABOUT ZEXEL

Zexel WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!