The emergence of smart contracts on blockchain platforms has ushered in a transformative era in digital transactions, offering significant advantages in terms of transparency, security, and efficiency. These self-executing contracts have gained widespread adoption across various sectors, from finance to supply chain management, fundamentally altering how agreements are enforced and transactions are conducted [1]. However, as the use of smart contracts continues to expand, it has brought to the forefront a pressing concern—the necessity to safeguard the privacy and confidentiality of sensitive data involved in smart contract execution.
This literature review embarks on a comprehensive exploration of the evolving landscape of data privacy within the realm of smart contracts. While the benefits of transparency and immutability offered by blockchain technology are well-acknowledged, the potential exposure of sensitive data poses a complex and multifaceted challenge [2]. Therefore, it becomes imperative to delve into this critical issue, not only to comprehend the current state of research but also to identify avenues for future investigation. The significance of this research lies in its capacity to bridge the gap between the transformative potential of smart contracts and the imperative need for robust data privacy measures. As organizations and individuals increasingly rely on blockchain-based systems for a wide range of transactions, the ramifications of data breaches or unauthorized access to sensitive information become more pronounced. Consequently, a nuanced understanding of data privacy within smart contracts is not merely an academic pursuit but a practical necessity.
1.1. SCOPE OF THE SURVEY
The survey encompasses various investigations and methodologies conducted globally by different authors, exploring different aspects of Smart Contract with Artificial Intelligence by means of natural language processing (NLP) and Machine Learning (ML) [30]–[38]. Upon reviewing multiple literature surveys, it is evident that the majority focus on utilizing Block Chain for security and privacy, while others leverage Artificial Intelligence techniques for similar purposes. To our knowledge, only a limited number of surveys explore the integration of Artificial Intelligence with Block Chain to ensure privacy, and among them, few provide a comprehensive explanation of the mechanisms for achieving privacy.
1.2. CONTRIBUTION
The survey we propose covers a time span from 2008 to 2023, spanning the last fifteen years. In phase-I, our methodological approach involves employing bibliometric analysis, facilitating a systematic examination of the research body. This method tracks trends, pinpoints key authors, and maps the field's evolution, delivering valuable insights into the knowledge state and emerging areas of interest. This enhances the review's comprehensiveness, establishing a robust foundation for steering future research directions. Moving into phase-II, we delve deeper into AI-based approaches for the identification and protection of sensitive data within smart contracts.
1.3. METHODOLOGY:
This section establishes a foundation for understanding the subsequent sections. This work is divided into 2 Major phases. As shown in flow diagram below
1.4. DATA COLLECTION:
To collect data, the primary source used was the 1Web of Science core collection corpus, globe’s largest database, 4dimensions, 1Google Scholar, 7Cross Ref, PubMed, 6Medline, and Scopus, which yielded 3800 articles published between 2008 and 2023. From these, Open access articles were collected first, then on this collection applied inclusion and exclusion criteria, to shortlist the huge collection, then documents were used for further analysis. We employed Data science tools on metadata for analysis and visualization of large number of research-articles. Determined the citations mean. Studied useful patterns of various type of articles, examine the rate of collaborative research-work, in order to find the productive authors, growth rate of publications, number of citations of an authors and co-authors etc.
Search Queries:
-
Privacy Protection of data in Smart Contracts from 2015 to 2023,
-
Machine learning for data privacy in blockchain from 2019 to 2023,
-
Data Privacy Mechanism in Smart contracts, issues and challenges from 2013 to 2023
-
Data Privacy in Smart Contracts from 2008 to 2023
Exploration of publication patterns and categories can illuminate the ever-evolving landscape of blockchain research. Esteemed contributors, establishments, nations, and periodicals were all engaged in this comprehensive bibliometric inquiry. Theoretical knowledge and prevailing research focal points within the blockchain realm encompassed keywords like 'blockchain', 'smart contract', 'privacy', 'bitcoin', 'security', 'Ethereum', 'sensitive data', and 'cryptography'. Employing a keyword-centered approach, a cluster analysis was executed to unveil emerging prospects and frontiers in the blockchain sphere. The results indicated the need for forthcoming research to prioritize areas including governance, blockchain technology, energy, machine learning, and smart home applications. As a result, this paper imparts valuable insights that offer scholars and practitioners a holistic grasp of the present landscape and future trajectories in blockchain research.
Table 1: Data collected apart from WoS from these sources also
Source
|
OpenAlex
|
Google Scholar
|
Crossref
|
Papers
|
1000
|
200
|
1000
|
Cites_Year
|
11427.63
|
4624
|
722.5
|
Cites_Paper
|
91.42
|
92.48
|
7.23
|
Cites_Author
|
32675.84
|
4671.34
|
3059.62
|
Papers_Author
|
380.71
|
54.71
|
520.6
|
h_index
|
149
|
76
|
38
|
g_index
|
244
|
126
|
73
|
hc_index
|
133
|
100
|
36
|
hI_norm
|
83
|
37
|
24
|
The provided values are metrics and indicators that offer insights into the research impact and productivity of a source such as OpenAlex, Google scholar and cross Ref.
1.5. RESEARCH CATEGORIES:
Research categories are subject classifications that aid in the categorization of publications into distinct domains or disciplines. These categories facilitate focused analysis on specific research areas and enable exploration of interdisciplinary connections. Scrutinizing citations according to the WoS (Web of Science) blockchain categories can provide insights into the progress of research within a designated timeframe and unveil areas that have garnered escalating interest over time [26]. Within the realm of WoS blockchain literature, there are approximately 98 distinct subject classifications, with 22 of these categories associated with the domain of Data Privacy.
The topmost eight subject categories were illustrated in Table 2, comprising Information and Computing Sciences (60492 articles), Commerce, Management, Tourism and Services (17817 articles), Human Society (15943 articles), Engineering (11399 articles), Law and Legal Studies (9059 articles), Education (4042 articles) and Biomedical and Clinical Sciences (3108 articles). The arrangement of subject classifications hinted at the emphasis placed on concerns within these areas in research endeavors. The quantity of publications within each category served as an indicator of the evolving trajectories of privacy mechanisms in blockchain research across diverse fields. Notably, the volume of publications in the domains of Information and Computing Sciences, as well as Commerce, Management, Tourism, and Services, witnessed a substantial surge post-2010. Conversely, the quantity of publications in Health Sciences, Biological Sciences, Biomedical and Clinical Science exhibited a more gradual increase.
This segment delineates the systematic method used for executing the literature review. Recent advancements in methodology by [ML10] are taken into account, assimilating their perspectives to enhance the precision of search strategies and criteria. Adhering to the PRISMA guidelines [ML11], the systematic review guarantees transparency and methodological thoroughness.
Table 2
Research Categories related to Smart Contract
Category
|
Publications
|
Information and Computing Sciences
|
60492
|
Commerce, Management, Tourism and Services
|
17817
|
Human Society
|
15943
|
Engineering
|
11399
|
Law and Legal Studies
|
9059
|
Language, Communication and Culture
|
7833
|
Creative Arts and Writing
|
5017
|
1.6. BACKGROUND
Leveraging blockchain technology, smart contracts have risen as a significant strategy for both automating and ensuring the security of digital transactions [3]. These self-executing contracts embody their agreement terms directly within code, deploying them on blockchain networks to execute predetermined actions automatically once specific conditions are met. Nevertheless, smart contracts frequently require the management of sensitive information [4], encompassing data like personally identifiable information (PII), fiscal records, and other sensitive details.
Consider a smart contract as a digital agreement where two or more parties establish predefined conditions for their agreements [5]. All parties have the ability to access and examine the contract, with any alterations strictly forbidden. As a result, trust between parties becomes less pivotal, as they can rely on the contract and the underlying blockchain technology. Smart contracts efficiently eliminate the necessity for an intermediary like an agent or broker, as illustrated in Fig. 3, by marginalizing their roles. For instance, envision this scenario: the need arises for a contract, spanning a range of purposes like real estate transactions (e.g., land, home, or farm acquisitions), loans, driver's license issuance, material procurement, without the desire to engage an intermediary such as a legal counsel or a real estate agent. In this context, smart contracts can be employed to define and enforce all requisite conditions. Smart contracts offer numerous advantages, including automation, trustworthiness, security, cost-effectiveness, efficiency, enhanced accessibility, auditability, cost savings, transparency, and swiftness [6–7].
Basically, smart contracts are written in Solidity programming language [8], follows a statically typed approach, necessitating the explicit declaration of variable types within the contract. It does not permit the presence of neither "undefined" nor "null" value, and any newly created variable is assigned a default value according to its type. The structure of Solidity smart contracts revolves around subcontracts, interfaces, and libraries, providing a well-organized framework. These agreements may cover a range of components, including state attributes, procedures, modifiers for functions, occurrences, structural categories, and enumeration types. Furthermore, Solidity agreements provide the option to inherit [9] from alternate contracts and the capacity to call upon other contracts to enhance their functionality.
Solidity serves as a specialized programming language specifically tailored for the implementation of smart contracts on the Ethereum blockchain. It has gained widespread adoption as an open-source language for blockchain and smart contract development. While initially designed for Ethereum, Solidity can also be utilized on other blockchain platforms like Hyperledger and many other.
1.7. PLATFORM FOR BLOCKCHAIN
Prominent blockchain platform which supports smart contracts are Ethereum, Stellar, Polkadot, Hyperledger, Neo, Bitcoin, Binance Smart Chain, Cardano, Solana, Avalanche, Tron, EOS.IO, Algorand, Avalanche and Tezos etc. There are numerous areas in which smart contracts are playing a major role few are listed here; Finance, Healthcare, Travel and Mobility, Real Estate, Publishing, Entertainment, Business and Corporate Governance etc [10–16].
Table 3
Comparative table showcasing various aspects of smart contract platforms in 2023.
Platform
|
Execution Environment
|
Supported Programming Languages
|
Turing Completeness
|
Permission Type
|
Transactions per Second
|
Consensus Mechanism
|
Ethereum
|
EVM (Ethereum Virtual Machine)
|
Solidity, Vyper
|
✓
|
Public
|
100,000(TPS)
|
PoS (Proof of Stake)
|
Binance Smart Chain
|
BSC Virtual Machine
|
Solidity, Vyper
|
✓
|
Public
|
3000(TPS)
|
DPoS (Delegated Proof of Stake)
|
Cardano
|
Plutus
|
PlutusScript
|
✓
|
Public
|
1000(TPS)
|
PoS (Proof of Stake)
|
Polkadot
|
Substrate
|
Ink!, Solidity
|
✓
|
Public
|
1000(TPS)
|
Nominated Proof of Stake (NPoS)
|
Solana
|
Solana Runtime
|
Rust, C, C++
|
✓
|
Public
|
65,000(TPS)
|
PoH (Proof of History), PoS (Proof of Stake)
|
Tezos
|
Michelson
|
LLL (Low-Level Lisp)
|
✓
|
Public
|
40(TPS)
|
PoS (Proof of Stake)
|
Algorand
|
AVM (Algorand Virtual Machine)
|
Transaction Execution Approval Language (TEAL), Python
|
𝛘
|
Public
|
6,000(TPS)
|
PPoS (Pure Proof of Stake)
|
Avalanche
|
Avalanche Platform
|
Go, JavaScript, Python, TypeScript
|
𝛘
|
Public
|
4,500(TPS)
|
Snow Consensus (depends on PoS)
|
Corda
|
Java Virtual Machine
|
Kotlin, Java
|
𝛘
|
Private
|
20,000(TPS)
|
RAFT (Consensus Algorithm)
|
EOS
|
EOS VM (EOS Virtual Machine)
|
C++
|
✓
|
Public
|
4,000(TPS)
|
DPoS (Delegated Proof of Stake)
|
Hyperledger Fabric
|
Docker
|
Go, JavaScript
|
✓
|
Consortium
|
Variable(TPS)
|
Practical Byzantine Fault Tolerance (PBFT)
|
Waves
|
Waves Platform
|
Ride
|
𝛘
|
Public
|
1,000(TPS)
|
PoS (Proof of Stake)
|
Stellar
|
Stellar Consensus Protocol
|
JavaScript
|
𝛘
|
Consortium
|
1,000(TPS)
|
SCP (Stellar Consensus Protocol)
|