Calling relationship investigation and application on Ethereum Blockchain System

Blockchain technologies are increasingly applied to build secure and efficient trustworthy software systems. As the core of the blockchain, large numbers of smart contracts are created and deployed on chain so that end users can call smart contracts to perform various operations in these systems. Such calling relationships between users and smart contracts construct the blockchain network. In order to analyze user behaviors and optimize blockchain systems, it is necessary to understand the calling relationships between blockchain end users and smart contracts in details. Therefore, this paper adopts the famous public blockchain platform Ethereum as a case to investigate the calling relationships. In particular, we conduct an empirical study on Ethereum based on more than 764 million smart contract calling records on 13.25 million blocks from August 2015 to September 2021. Four research questions are proposed and answered: (1) We draw an overall picture of calling behaviors on Ethereum by using four overview indices, which are new user, active user, user retention and user attrition. (2) We categorize users by constructing RFM models based on three user indices, which are last call interval, call frequency and service charge. (3) We categorize smart contracts by constructing BCG matrices with two contract indices including market share and gas growth. (4) We conduct ETH price prediction and smart contract recommendation by employing the analysis results. Our observations and implications aim to provide some inspirations on further researches and applications for enhancing the activeness of Ethereum and improving the efficiency of smart contracts.