Abstract:Smart contract has been the core of blockchain systems and other blockchain-based systems since Blockchain 2.0. Various operations on blockchain are performed through the invocation and execution of smart contracts. This leads to extensive combinations between blockchain, smart contract, Internet of Things (IoT) and Cyber-Physical System (CPS) applications, and then many blockchain-based IoT or CPS applications emerge to provide multiple benefits to the economy and society. In this case, obtaining a better understanding of smart contracts will contribute to the easier operation, higher efficiency and stronger security of those blockchain-based systems and applications. Many existing studies on smart contract analysis are based on similarity calculation and smart contract classification. However, smart contract is a piece of code with special characteristics and most of smart contracts are stored without any category labels, which leads to difficulties of smart contract classification. As the back end of a blockchain-based Decentralized Application (DApp) is one or several smart contracts, DApps with labeled categories and open source codes are applied to achieve a supervised smart contract classification. A three-phase approach is proposed to categorize DApps based on various data features. In this approach, 5,659 DApps with smart contract source codes and pre-tagged categories are first obtained based on massive collected DApps and smart contracts from Ethereum, State of the DApps and DappRadar. Then feature extraction and construction methods are designed to form multi-feature vectors that could present the major characteristics of DApps. Finally, a fused classification model consisting of KNN, XGBoost and random forests is applied to the multi-feature vectors of all DApps for performing DApp classification. The experimental results show that the method is effective. In addition, some positive correlations between feature variables and categories, as well as several user behavior patterns of DApp calls, are found in this paper.

Data and Behavior Analysis of Blockchain-based DApp

A First Look at Blockchain-based Decentralized Applications.

An Empirical Study of Blockchain-based Decentralized Applications

Portrait of decentralized application users: an overview based on large-scale Ethereum data

Anomaly Analysis of Blockchain-based Decentralized Applications of Transportation

DAppHunter: Identifying Inconsistent Behaviors of Blockchain-based Decentralized Applications.

Applying blockchain-based method to smart contract classification for CPS applications

BlockChain and Decentralized Apps

A Preliminary Analysis of Software Metrics in Decentralised Applications

DApps Ecosystems: Mapping the Network Structure of Smart Contract Interactions

A Behavior-Aware Profiling Of Smart Contracts

Recent Advances of Blockchain and Its Applications

A Comparative Examination of Network and Contract-Based Blockchain Storage Solutions for Decentralized Applications

A survey of application research based on blockchain smart contract

Toward Open Data Blockchain Analytics: A Bitcoin Perspective

Ethereum DeFi Apps in the Wild: Profiling and Implications

A Survey of Blockchain Applications in Different Domains

A Statistical Study Based on the Top 50 Blockchain Companies Analyzing the Current Development of Blockchain Applications

Security and privacy protection technologies in securing blockchain applications

Advancing Consumer Adoption of Blockchain Applications

BLOCKCHAIN TECHNOLOGIES AND APPLICATIONS